Anti-CD28×anti-TROP2 antibody

Abstract

This specification provides novel anti-CD28 × anti-TROP2 antibodies and methods for using such antibodies for the treatment of TROP2-related cancers. The target anti-CD28 × anti-TROP2 antibodies can agonist-conjugate CD28 costimulatory molecules on T cells and TROP2 on tumor cells. Thus, such antibodies selectively enhance antitumor activity at tumor sites while minimizing peripheral toxicity. The target antibodies provided herein are particularly useful in combination with other anticancer therapies, for example, bispecific antibodies for the treatment of TROP2-related cancers.

Description

[Technical Field]

[0001] Claim of priority This application claims priority and interest to U.S. Provisional Application No. 63 / 496,372 filed on 14 April 2023 and U.S. Provisional Application No. 63 / 593,942 filed on 27 October 2023, the contents of which are incorporated herein by reference in their entirety. [Background technology]

[0002] Tumor-associated calcium signaling transducer 2 (TROP2) is a transmembrane protein overexpressed in various malignancies and is an oncogene associated with the development, invasion, and metastasis of malignancies. Exemplary cancers that overexpress TROP2 include endometrial cancer, pancreatic cancer, colon cancer, esophageal cancer, cholangiocarcinoma, oral cancer, and glioma. Because TROP2 is overexpressed in a variety of tumors, it is a candidate for targeted therapy development.

[0003] Antibody-based therapies have been used to successfully treat a variety of diseases, including cancer. An increasingly widespread approach being explored is the manipulation of a single immunoglobulin molecule that co-engages with two different antigens. Such alternative antibody forms that engage with two different antigens are often referred to as bispecific antibodies. One particular approach for bispecific antibodies is that the bispecific antibody is CD3 + The goal is to design a first binding domain that engages with CD3 to redirect T cells and destroy cancer cells, and a second binding domain that engages with an antigen associated with or upregulated on cancer cells (e.g., TROP2).

[0004] However, TILs lose their cytotoxic capacity over time due to upregulation of inhibitory immune checkpoints. While checkpoint blockade has demonstrated increased clinical response rates compared to other treatment options, many patients still fail to achieve a response to checkpoint blockade. Engagement of costimulatory receptors on TILs could provide a positive signal that could overcome the negative signaling of immune checkpoints. Preclinical and clinical studies of agonist costimulatory receptor antibodies have indeed demonstrated that costimulatory receptor agonism can lead to superior antitumor responses by activating T cells to attack tumor cells.

[0005] Improving antitumor activity by specifically destroying tumor cells while minimizing peripheral toxicity is also important for cancer therapy. In this context, it is crucial that costimulatory signals are delivered only to T cells in the presence of target tumor cells. However, costimulatory receptor agonism with single-specific full-length antibodies is likely not discriminatory with respect to TILs, peripheral T cells, or autoantigen-reactive T cells that contribute to autoimmune toxicity.

[0006] While immunotherapies targeting TROP2 are being explored, novel immune response-enhancing compositions are still needed to treat TROP2-related cancers. [Overview of the project]

[0007] This specification provides novel anti-CD28 × anti-TROP2 antibodies and methods for using such antibodies for the treatment of TROP2-related cancers. The target anti-CD28 × anti-TROP2 antibodies can agonist-conjugate CD28 costimulatory molecules on T cells and TROP2 on tumor cells. Thus, such antibodies selectively enhance antitumor activity at tumor sites while minimizing peripheral toxicity. The target antibodies provided herein are particularly useful in combination with other anticancer therapies, for example, bispecific antibodies for the treatment of TROP2-related cancers.

[0008] In a first aspect, the herein provides an anti-CD28 × anti-TROP2 heterodimer antibody comprising a) a first monomer, b) a second monomer, and c) a light chain. The first monomer comprises i) a single-stranded variable fragment (scFv) and ii) a first Fc domain, the scFv being covalently bound to the N-terminus of the first Fc domain using a domain linker. The second monomer comprises VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, where VH1 is the first variable heavy domain and CH2-CH3 is the second Fc domain. The light chain comprises VL1-CL from the N-terminus to the C-terminus, where VL1 is the first variable light domain and CL is the constant light domain. The scFv comprises a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2). VH1 and VL1 together form a first antigen-binding domain (ABD), and VH2 and VL2 together form a second ABD. Furthermore, one of the first and second ABDs is a CD28-binding domain, and the other is a tumor-associated calcium signaling transducer 2 (TROP2)-binding domain.

[0009] In some embodiments, the scFv includes a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

[0010] In exemplary embodiments, the first ABD is a TROP2 binding domain, and the second ABD is a CD28 binding domain. In some embodiments, VH1 and VL1 are VH and VL or variants thereof of any of the TROP2 binding domains in Figures 23-26 and 52-56. In some embodiments, VH2 and VL2 are selected from one of the following: 1) VH and VL or variants thereof of any of the CD28 binding domains in Figures 16, 19, and 22, or 2) (i) VH or a variant thereof in Figures 16, 17, or 34; and (ii) VL or a variant thereof in Figures 16, 18, or 36.

[0011] In some embodiments, the first Fc domain and the second Fc domain are variant Fc domains.

[0012] In some embodiments, the first and second Fc domains include a set of heterodimerized scuba riants selected from the following heterodimerized variants:S364K / E357Q:L368D / K370S;S364K:L368D / K370S;S364K:L368E / K370S;D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, numbered according to EU numbering. In some embodiments, the first and second Fc domains include the heterodimerized scuba riant S364K / E357Q:L368D / K370S, numbered according to EU numbering.

[0013] In certain embodiments, the first and second Fc domains each include one or more attenuation variants. In exemplary embodiments, the one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, the numbering following EU numbering.

[0014] In some embodiments, one of the first or second monomer further comprises one or more pI variants. In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer comprises pI variants N208D / Q295E / N384D / Q418E / N421D, the numbering following EU numbering.

[0015] In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, and the first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, with the numbering following EU numbering. In exemplary embodiments, the first and second variant Fc domains each further contain the amino acid variant 428L / 434S.

[0016] In some embodiments, the scFv linker is GKPGSGKPGSGKPGSGKPGS (Sequence ID 24).

[0017] In another embodiment, provided herein is an anti-CD28 × anti-TROP2 heterodimer antibody comprising a) a first monomer, b) a second monomer, c) a first light chain, and d) a second light chain. The first monomer comprises VH1-CH1-first domain linker-scFv-second domain linker-CH2-CH3 from N-terminus to C-terminus, where VH1 is the first variable heavy domain and CH2-CH3 is the first Fc domain. The second monomer comprises VH1-CH1-hinge-CH2-CH3 from N-terminus to C-terminus, where CH2-CH3 is the second Fc domain. The first and second light chains each comprise VL1-CL from N-terminus to C-terminus, where VL1 is the first variable light domain and CL is the constant light domain. scFv comprises a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2). The first monomer VH1 and the first light chain VL1, and the second monomer VH1 and the second light chain VL1, each form a first antigen-binding domain (ABD), while VH2 and VL2 form a second ABD. Furthermore, one of the first and second ABDs is a CD28-binding domain, and the other of the first and second ABDs is a tumor-associated calcium signaling transducer 2 (TROP2)-binding domain.

[0018] In some embodiments, the scFv includes a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

[0019] In exemplary embodiments, the first ABD is a TROP2 binding domain, and the second ABD is a CD28 binding domain. In some embodiments, VH1 and VL1 are selected from one of the following: 1) VH and VL or variants thereof of any of the TROP2 binding domains in Figures 23-26 and 52-56. In some embodiments, VH2 and VL2 are selected from one of the following: 1) VH and VL or variants thereof of any of the CD28 binding domains in Figures 16, 19, and 22, or 2) (i) VH or a variant thereof in Figures 16, 17, or 34; and (ii) VL or a variant thereof in Figures 16, 18, or 36.

[0020] In some embodiments, the first Fc domain and the second Fc domain are variant Fc domains.

[0021] In some embodiments, the first and second Fc domains include a set of heterodimerized scuba riants selected from the following heterodimerized variants:S364K / E357Q:L368D / K370S;S364K:L368D / K370S;S364K:L368E / K370S;D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, numbered according to EU numbering. In some embodiments, the first and second Fc domains include the heterodimerized scuba riant S364K / E357Q:L368D / K370S, numbered according to EU numbering.

[0022] In certain embodiments, the first and second Fc domains each include one or more attenuation variants. In exemplary embodiments, the one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, the numbering following EU numbering.

[0023] In some embodiments, one of the first or second monomers further comprises one or more pI variants. In some embodiments, the CH1 - hinge - CH2 - CH3 of the second monomer comprises the pI variant N208D / Q295E / N384D / Q418E / N421D, and the numbering follows the EU numbering.

[0024] In some embodiments, the CH1 - hinge - CH2 - CH3 of the second monomer comprises the amino acid variants E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, and the first Fc domain comprises the amino acid variants E233P / L234V / L235A / G236del / S267K / S364K / E357Q, and the numbering follows the EU numbering. In an exemplary embodiment, the first and second variant Fc domains each further comprise the amino acid variant 428L / 434S.

[0025] In some embodiments, the scFv linker is GKPGSGKPGSGKPGSGKPGSGKPGS (SEQ ID NO: 24).

[0026] In another embodiment, provided herein is an anti-CD28 × anti-TROP2 heterodimer antibody comprising a) a first monomer, b) a second monomer, c) a first light chain, and d) a second light chain. The first monomer comprises VH1-CH1-hinge-CH2-CH3-domain linker-scFv from N-terminus to C-terminus, where VH1 is the first variable weight domain and CH2-CH3 is the first Fc domain. The second monomer comprises VH1-CH1-hinge-CH2-CH3 from N-terminus to C-terminus, where VH1 is the first variable weight domain and CH2-CH3 is the second Fc domain. The first and second light chains each comprise VL1-CL from N-terminus to C-terminus, where VL1 is the first variable light domain and CL is the constant light domain. scFv comprises a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2). The VH1 of the first monomer and the VL1 of the first light chain, and the VH1 of the second monomer and the VL1 of the second light chain, each form a first antigen-binding domain (ABD), while VH2 and VL2 form a second ABD. Furthermore, one of the first and second ABDs binds to human CD28, and the other of the first and second ABDs binds to TROP2.

[0027] In some embodiments, the scFv includes a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

[0028] In exemplary embodiments, the first ABD is a TROP2 binding domain, and the second ABD is a CD28 binding domain. In some embodiments, VH1 and VL1 are selected from one of the following: 1) VH and VL or variants thereof of any of the TROP2 binding domains in Figures 23-26 and 52-56. In some embodiments, VH2 and VL2 are selected from one of the following: 1) VH and VL or variants thereof of any of the CD28 binding domains in Figures 16, 19, and 22, or 2) (i) VH or a variant thereof in Figures 16, 17, or 34; and (ii) VL or a variant thereof in Figures 16, 18, or 36.

[0029] In some embodiments, the first Fc domain and the second Fc domain are variant Fc domains.

[0030] In some embodiments, the first and second Fc domains include a set of heterodimerized scuba riants selected from the following heterodimerized variants:S364K / E357Q:L368D / K370S;S364K:L368D / K370S;S364K:L368E / K370S;D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, numbered according to EU numbering. In some embodiments, the first and second Fc domains include the heterodimerized scuba riant S364K / E357Q:L368D / K370S, numbered according to EU numbering.

[0031] In certain embodiments, the first and second Fc domains each include one or more attenuation variants. In exemplary embodiments, the one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, the numbering following EU numbering.

[0032] In some embodiments, one of the first or second monomer further comprises one or more pI variants. In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer comprises pI variants N208D / Q295E / N384D / Q418E / N421D, the numbering following EU numbering.

[0033] In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, and the first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, with the numbering following EU numbering. In exemplary embodiments, the first and second variant Fc domains each further contain the amino acid variant 428L / 434S.

[0034] In some embodiments, the scFv linker is GKPGSGKPGSGKPGSGKPGS (Sequence ID 24).

[0035] In another embodiment, provided herein is a bispecific antibody comprising a TROP2-binding domain comprising a)i) a first variable heavy domain (VH1) and ii) a first variable light domain (VL1), and an anti-CD28-binding domain comprising b)i) a second variable heavy domain (VH2) and ii) a second variable light domain (VL2). In some embodiments, VH1 and VL1 are selected from one of the following: 1) VH and VL or variants thereof of any of the TROP2-binding domains in Figures 23-26 and 52-56. In some embodiments, VH2 and VL2 are selected from one of the following: 1) VH and VL or variants thereof of any of the CD28-binding domains in Figures 16, 19, and 22, or 2) (i) VH or a variant thereof in Figures 16, 17, or 34; and (ii) VL or a variant thereof in Figures 16, 18, or 36.

[0036] In some embodiments of the bispecific antibody, the first Fc domain and the second Fc domain are variant Fc domains, respectively.

[0037] In some embodiments of the bispecific antibody, the first and second Fc domains include a set of heterodimerized scuba riants selected from the following heterodimerized variants:S364K / E357Q:L368D / K370S;S364K:L368D / K370S;S364K:L368E / K370S;D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, numbered according to EU numbering. In some embodiments, the first and second Fc domains include the heterodimerized scuba riant S364K / E357Q:L368D / K370S, numbered according to EU numbering.

[0038] In certain embodiments, the first and second Fc domains each include one or more attenuation variants. In exemplary embodiments, the one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, the numbering following EU numbering.

[0039] In some embodiments, one of the first or second monomer further comprises one or more pI variants. In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer comprises pI variants N208D / Q295E / N384D / Q418E / N421D, the numbering following EU numbering.

[0040] In one embodiment, the herein provides a polyvalent antibody comprising a first monomer, a second monomer, a first common light chain, and a second common light chain. The first monomer comprises VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, where VH1 is the first variable weight domain and CH2-CH3 is the first Fc domain. The second monomer comprises VH2-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, where VH2 is the second variable weight domain and CH2-CH3 is the second Fc domain. The first and second common light chains each comprise VL-CL, where VL is the variable domain and CL is the constant light domain. The VH1 and VL of the first common light chain form the first antigen-binding domain (ABD), and the VH2 and VL of the second common light chain together form the second ABD, with the first and second common light chains having the same amino acid sequence. Furthermore, one of the first and second ABDs is a CD28-binding domain, and the other of the first and second ABDs is a tumor-associated calcium signaling transducer 2 (TROP2)-binding domain.

[0041] In some embodiments, the first ABD is a TROP2 binding domain, and the second ABD is a CD28 binding domain. In some embodiments, VH1 is a variable weight domain as shown in Figures 36 or 52-56, VH2 is a variable weight domain as shown in Figure 34, and the VL of the first and second common light chains is the 1F11-1A3.315[TROP2]_L1 variable light domain (see Figure 36A).

[0042] In some embodiments, the first Fc domain and the second Fc domain are variant Fc domains.

[0043] In some embodiments, the first and second Fc domains include a set of heterodimerized scuba riants selected from the following heterodimerized variants:S364K / E357Q:L368D / K370S;S364K:L368D / K370S;S364K:L368E / K370S;D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, numbered according to EU numbering. In some embodiments, the first and second Fc domains include the heterodimerized scuba riant S364K / E357Q:L368D / K370S, numbered according to EU numbering.

[0044] In certain embodiments, the first and second Fc domains each include one or more attenuation variants. In exemplary embodiments, the one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, the numbering following EU numbering.

[0045] In some embodiments, one of the first or second monomer further comprises one or more pI variants. In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer comprises pI variants N208D / Q295E / N384D / Q418E / N421D, the numbering following EU numbering.

[0046] In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, and the first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, with the numbering following EU numbering. In exemplary embodiments, the first and second variant Fc domains each further contain the amino acid variant 428L / 434S.

[0047] In another embodiment, the herein provides a polyvalent antibody comprising a first monomer, a second monomer, a first common light chain, a second common light chain, and a third common light chain. The first monomer comprises VH1-CH1-linker-VH1-CH1-hinge-CH2-CH3 from N-terminus to C-terminus, where VH1 is the first variable weight domain and CH2-CH3 is the first Fc domain. The second monomer comprises VH2-CH1-hinge-CH2-CH3 from N-terminus to C-terminus, where VH2 is the second variable weight domain and CH2-CH3 is the second Fc domain. The first common light chain, the second common light chain, and the third common light chain each comprise VL-CL, where VL is the variable domain and CL is the constant light domain. Each of the first monomer VH1 pairs with either the first common light chain or the second common light chain VL to form two first antigen-binding domains (ABDs), while the third common light chain VH2 and VL form a second ABD, and the first, second, and third common light chains each have the same amino acid sequence. Furthermore, each of the first ABDs is a CD28-binding domain and the second ABD is a tumor-associated calcium signaling transducer 2 (TROP2)-binding domain, or each of the first ABDs is a TROP2-binding domain and the second ABD is a CD28-binding domain.

[0048] In some embodiments, the first ABD is a TROP2 binding domain, and the second ABD is a CD28 binding domain. In some embodiments, VH1 is a variable weight domain as shown in Figures 36 or 52-56, VH2 is a variable weight domain as shown in Figure 34, and the VL of the first, second, and third common light chains is the 1F11-1A3.315[TROP2]_L1 variable light domain (see Figure 36A).

[0049] In some embodiments, the first Fc domain and the second Fc domain are variant Fc domains.

[0050] In some embodiments, the first and second Fc domains include a set of heterodimerized scuba riants selected from the following heterodimerized variants:S364K / E357Q:L368D / K370S;S364K:L368D / K370S;S364K:L368E / K370S;D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, numbered according to EU numbering. In some embodiments, the first and second Fc domains include the heterodimerized scuba riant S364K / E357Q:L368D / K370S, numbered according to EU numbering.

[0051] In certain embodiments, the first and second Fc domains each include one or more attenuation variants. In exemplary embodiments, the one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, the numbering following EU numbering.

[0052] In some embodiments, one of the first or second monomer further comprises one or more pI variants. In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer comprises pI variants N208D / Q295E / N384D / Q418E / N421D, the numbering following EU numbering.

[0053] In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, and the first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, with the numbering following EU numbering. In exemplary embodiments, the first and second variant Fc domains each further contain the amino acid variant 428L / 434S.

[0054] Furthermore, this specification also provides nucleic acid compositions comprising nucleic acids encoding antibodies described herein, expression vector compositions comprising such nucleic acids, host cells for producing antibodies comprising the expression vector compositions, and methods for producing antibodies.

[0055] In another embodiment, the foregoing provides a method for treating TROP2-associated cancer in a patient requiring treatment for TROP2-associated cancer, the method comprising administering to the patient an anti-CD28 × anti-TROP2 bispecific antibody as described herein.

[0056] In another embodiment, the foregoing provides a method for treating TROP2-associated cancer in a patient requiring treatment for TROP2-associated cancer, comprising administering to the patient an anti-CD28 × anti-TROP2 bispecific antibody and an anti-CD3 × anti-TROP2 bispecific antibody as described herein.

[0057] In one embodiment, the TROP-2 binding construct provided herein includes a variable heavy domain and a variable light domain selected from: 1) VH and VL or variants thereof of either the TROP2 binding domains VH and VL shown in Figures 23-26 and 52-56; and 2) VH or a variant thereof shown in Figure 36, and 1F11-1A3.315L1VL (see Figure 36A) or a variant thereof. [Brief explanation of the drawing]

[0058] [Figure 1A] The sequences of human, mouse, and cynomolgus monkey CD28 are shown. Such CD28 is useful for developing cross-reactive CD28 antigen-binding domains to facilitate clinical development. [Figure 1B] The sequences of human, mouse, and cynomolgus monkey CD28 are shown. Such CD28 is useful for developing cross-reactive CD28 antigen-binding domains to facilitate clinical development. [Figure 2]The sequences of human, mouse, and cynomolgus monkey TROP2 are shown. Such TROP2s are useful for developing cross-reactive TROP2 antigen-binding domains to facilitate clinical development. [Figure 3A] This shows useful pairs of heterodimerized variant sets (including skew and pI variants). [Figure 3B] This shows useful pairs of heterodimerized variant sets (including skew and pI variants). [Figure 3C] This shows useful pairs of heterodimerized variant sets (including skew and pI variants). [Figure 3D] This shows useful pairs of heterodimerized variant sets (including skew and pI variants). [Figure 3E] This shows useful pairs of heterodimerized variant sets (including skew and pI variants). [Figure 3F] Useful pairs of heterodimerized variant sets (including skew and pI variants) are shown. Figure 3F shows variants for which there is no variant corresponding to "monomer 2". Such variants are pI variants and can be used alone for either monomer of a bispecific antibody (e.g., TROP2×CD28 bsAb), or may be included in the non-scFv side in a form that utilizes scFv as a component, for example, and a suitable charged scFv linker can be used for the second monomer that utilizes scFv as the CD28 binding domain. Suitable charged linkers are shown in Figure 6. [Figure 4] A list of isosteric variant antibody constant regions and their respective substitutions is shown. pI_(-) indicates a lower pI variant, while pI_(+) indicates a higher pI variant. These variants can be combined arbitrarily and independently with other variants, including the heterodimerized variants outlined herein. [Figure 5]The document exhibits useful attenuation variants (also referred to as "knockout" or "KO" variants) that reduce FcγR binding. In some embodiments, such attenuation variants are present in the Fc domains of both monomers of the target antibody described herein. In other embodiments, the attenuation variant is present in only one variant's Fc domain. [Figure 6A] As described herein, we present a number of charged scFv linkers used to increase or decrease the pI of a heterodimer bispecific antibody (e.g., TROP2×CD28 bsAb) that utilizes one or more scFv components. [Figure 6B] As described herein, numerous charged scFv linkers are used to increase or decrease the pI of heterodimer bispecific antibodies (e.g., TROP2×CD28 bsAb) that utilize one or more scFv components. (+H) positive linkers are particularly utilized herein, especially in conjunction with the anti-CD28 VL and VH sequences shown herein. A single-charged, single-prior-art scFv linker is referred to as "Whitlow" from Whitlow et al., Protein Engineering 6(8):989-995 (1993). It should be noted that this linker was used to reduce aggregation and improve proteolytic stability in scFv. Such charged scFv linkers may be used in any of the target antibody forms disclosed herein that include scFv (e.g., 1+1 Fab-scFv-Fc and 2+1 Fab2-scFv-Fc forms). [Figure 7] Numerous exemplary domain linkers are shown. In some embodiments, these linkers are used to link a single-stranded Fv to an Fc chain. In some embodiments, these linkers can be combined in any orientation. For example, the GGGGS linker (SEQ ID NO: 37) can be combined with a “bottom half-hinge” linker at the N-terminus or C-terminus. [Figure 8]This invention presents a bispecific antibody platform particularly useful for TROP2×CD28 bsAb. While the platform is shown in relation to the 1+1 Fab-scFv-Fc format, it can also be applied to the use of other bispecific antibody formats. [Figure 9] This specification shows various heterodimer scuba rian amino acid substitutions that can be used with the heterodimer antibodies described herein. [Figure 10A] Based on human IgG1, the sequences of several useful heterodimer TROP2×CD28 bsAb skeletons, which do not contain cytokine sequences, are shown. [Figure 10B] Based on human IgG1, the sequences of several useful heterodimer TROP2×CD28 bsAb skeletons, which do not contain cytokine sequences, are shown. [Figure 10C] Based on human IgG1, the sequences of several useful heterodimer TROP2×CD28 bsAb skeletons, which do not contain cytokine sequences, are shown. [Figure 10D]Based on human IgG1, the sequences of several useful heterodimer TROP2×CD28 bsAb skeletons, which do not contain cytokine sequences, are shown. Heterodimer Fc skeleton 1, based on human IgG1 (356E / 358M allotype), includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant on both chains. The heterodimer Fc skeleton 2 is based on human IgG1 (356E / 358M allotype) and includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant on both chains. The heterodimer Fc skeleton 3 is based on human IgG1 (356E / 358M allotype) and includes the L368E / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant on both chains. The heterodimer Fc skeleton 4 is based on human IgG1 (356E / 358M allotype) and includes the K360E / Q362E / T411E scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the D401K scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant on both chains.The heterodimer Fc skeleton 5 is based on human IgG1 (356D / 358L allotype) and includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant on both chains. The heterodimer Fc skeleton 6 is based on human IgG1 (356E / 358M allotype) and includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant and the N297A variant that removes glycosylation on both chains. The heterodimer Fc skeleton 7 is based on human IgG1 (356E / 358M allotype) and includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K decrement variant and the N297S variant that removes glycosylation on both chains. The heterodimer Fc skeleton 8 is based on human IgG4 and includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variants on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the S228P (according to EU numbering, S241P in Kabat) variant which reduces Fab arm exchange (as known in the art) on both chains. The heterodimer Fc skeleton 9 is based on human IgG2 and includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variants on the first heterodimer Fc chain, and the S364K / E357Q scuba riant on the second heterodimer Fc chain.The heterodimer Fc skeleton 10 is based on human IgG2 and includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the S267K attenuation variant on both chains. The heterodimer Fc skeleton 11 is based on human IgG1 (356E / 358M allotype) and includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant and the M428L / N434S Xtend variant on both chains. The heterodimer Fc skeleton 12 is based on human IgG1 (356E / 358M allotype) and includes the L368D / K370S scuba rianto on the first heterodimer Fc chain, the S364K / E357Q scuba rianto and P217R / P229R / N276K pI variant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant on both chains. The heterodimer Fc skeleton 13, based on human IgG1 (356D / 358L allotype), includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant and M428L / N434S Xtend variant on both chains. The heterodimer Fc skeleton 14, based on human IgG1 (356E / 358M allotype), includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant and the M428L / N434A Xtend variant on both chains.The heterodimer Fc skeleton 15 is based on human IgG1 (356D / 358L allotype) and includes the L368D / K370S scuba riant and Q295E / N384D / Q418E / N421D pI variant on the first heterodimer Fc chain, the S364K / E357Q scuba riant on the second heterodimer Fc chain, and the E233P / L234V / L235A / G236del / S267K attenuation variant and the M428L / N434A Xtend variant on both chains. Sequences that are 90, 95, 98, and 99% identical (as defined herein) to the described sequences, and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 or more additional amino acid substitutions (as understood by those skilled in the art, compared to parent human IgG1 (or IgG2 or IgG4 depending on the skeleton), compared to the “parent” in the figure which already contains numerous amino acid modifications) are included in each of these skeletons. That is, the described skeletons may contain additional amino acid modifications (usually amino acid substitutions) in addition to or as alternatives to the skew, pI, and attenuation variants contained within the skeleton in this figure. Furthermore, the skeletons shown herein may include deletions of C-terminal glycine (K446_) and / or lysine (K447_). C-terminal glycine and / or lysine deletions may be intentionally manipulated to reduce heterogeneity or in relation to a predetermined bispecific form such as the mAb-scFv form. Furthermore, C-terminal glycine and / or lysine deletions may occur naturally, for example, during production and storage. [Figure 11] Exemplary sequences of the heterodimer TROP2×CD28 bsAb skeleton for use in the 2+1 mAb-scFv form are shown. The form shown here is based on the heterodimer Fc skeleton 1 shown in Figure 10, but further including G446_ on monomer 1(-) and G446_ / K447_ on monomer 2(+). It should be noted that any of the additional skeletons shown in Figure 10 can be applied to use in the 2+1 mAb-scFv form, whether or not they include K447_ on one or both strands. It should be noted that these sequences may further include the M428L / N434S variant. [Figure 12]The sequence of "CH1" used in an embodiment of TROP2×CD28 bsAb is shown. [Figure 13] This shows the arrangement of "hinge" used in the embodiment of TROP2×CD28 bsAb. [Figure 14] This shows the constant domain of the congeneral light chain used in TROP2×CD28 bsAb, which utilizes the Fab-binding domain. [Figure 15A] The present invention presents a bispecific form. It presents a "1+1 Fab-scFv-Fc" form having a first Fab arm that binds to a first antigen and a second scFv arm that binds to a second antigen. The 1+1 Fab-scFv-Fc form comprises a first monomer containing a first heavy chain variable region (VH1) covalently bonded (optionally via a linker) to the N-terminus of a first heterodimer Fc skeleton, a second monomer containing a single-stranded Fv covalently bonded (optionally via a linker) to the N-terminus of a second corresponding heterodimer Fc skeleton, and a third monomer containing a light chain variable region covalently bonded to a light chain constant domain (the light chain variable region is complementary to VH1). [Figure 15B] The present invention presents a bispecific form. It presents a "2+1 Fab2-scFv-Fc" form having a first Fab arm and a second Fab-scFv arm, where Fab binds to the first antigen and scFv binds to the second antigen. The 2+1 Fab2-scFv-Fc form comprises a first monomer containing a first heavy chain variable region (VH1) covalently bonded (optionally via a linker) to the N-terminus of a first heterodimer Fc skeleton, a second monomer containing VH1 covalently bonded (optionally via a linker) to a single-stranded Fv covalently bonded (optionally via a linker) to the N-terminus of a second corresponding heterodimer Fc skeleton, and a third monomer containing a light chain variable region covalently bonded to a light chain constant domain (the light chain variable region is complementary to VH1). [Figure 15C]The present invention presents a bispecific form. It presents a "1+1 common light chain" or "1+1 CLC" form having a first Fc containing a first Fab arm that binds to a first antigen and a second Fc containing a second Fab arm that binds to a second antigen. The 1+1 CLC form comprises a first monomer containing VH1-CH1-hinge-CH2-CH3, a second monomer containing VH2-CH1-hinge-CH2-CH3, and a third monomer containing VL-CL. VL pairs with VH1 to form a binding domain having a first antigen-binding specificity, and VL pairs with VH2 to form a binding domain having a second antigen-binding specificity. [Figure 15D] The present invention presents a bispecific form. It presents a "2+1 common light chain" or "2+1 CLC" form having a first Fc containing two Fab arms that bind to a first antigen and a second Fc containing one Fab arm that binds to a second antigen. The 2+1 CLC form comprises a first monomer containing VH1-CH1-linker-VH1-CH1-hinge-CH2-CH3, a second monomer containing VH2-CH1-hinge-CH2-CH3, and a third monomer containing VL-CL. VL pairs with the first and second VH1 to form a binding domain having a first antigen-binding specificity, and VL pairs with VH2 to form a binding domain having a second antigen-binding specificity. [Figure 15E] The present invention presents a bispecific form. It presents a "2+1 mAb-scFv" form having a first Fc containing an N-terminal Fab arm that binds to a first antigen, and a second Fc containing an N-terminal Fab arm that binds to a first antigen and a C-terminal scFv that binds to a second antigen. The 2+1 mAb-scFv form comprises a first monomer containing VH1-CH1-hinge-CH2-CH3, a second monomer containing VH1-CH1-hinge-CH2-CH3-scFv, and a third monomer containing VL-CL. VL pairs with the first and second VH1 to form a binding domain having binding specificity for the first antigen. [Figure 15F] This shows the dual-specific form of the present invention. It is a dual scFv, an additional dual-specific form. [Figure 15G]The present invention presents a dual-specific form. An additional dual-specific form is a 1-arm scFv-mAb. [Figure 15H] The present invention presents a bispecific form. An additional bispecific form is scFv-mAb. [Figure 15I] This presents a bispecific form of the present invention. It is a bispecific mAb of an additional bispecific form. [Figure 15J] The present invention presents a dual-specific form. An additional dual-specific form is a 1-arm central-scFv. [Figure 15K] The present invention presents a bispecific form. An additional bispecific form is mAb-Fv. [Figure 15L] The present invention presents a bispecific form. The additional bispecific form is the central-Fv. [Figure 15M] This presents a dual-specific form of the present invention. It is a trident of additional dual-specific forms. [Figure 15N] The present invention exhibits a bispecific form, specifically the 2+1 Fab2-Fc×scFv-Fc form (also known as the "stacked bottle opener" form). This form comprises a first monomer containing a VH1-CH1-domain linker-VH1-CH1-hinge-CH2-CH3, a second monomer containing scFv covalently bound to the Fc domain (CH2-CH3), and a third monomer containing VL-CL. VL pairs with the first and second VH1 of the first monomer to form a binding domain having binding specificity for the first antigen, while scFv has binding specificity for the second antigen. [Figure 16]The variable heavy chain and variable light chain sequences of 1A7, an exemplary phage-derived CD28-binding domain, are shown, as well as the sequence of XENP28428, an anti-CD28 mAb based on the IgG1 scaffold with 1A7 and the E233P / L234V / L235A / G236del / S267K attenuation variant. CDRs are indicated by underlines, and slashes indicate the boundary(s) between the variable region and the constant domain. As is the case with all sequences described herein that contain CDRs, the precise location of the CDR may vary slightly depending on the numbering used, as shown in Table 2. Therefore, this specification includes not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems. Furthermore, for all sequences in the figures, these VH and VL sequences may be used in either scFv or Fab format. [Figure 17A] The sequence of the affinity-optimized variable weight domain derived from anti-CD28 clone 1A7 is shown. [Figure 17B] The sequence of the affinity-optimized variable weight domain derived from anti-CD28 clone 1A7 is shown. [Figure 17C] The sequence of the affinity-optimized variable weight domain derived from anti-CD28 clone 1A7 is shown. [Figure 17D] The sequence of the affinity-optimized variable weight domain derived from anti-CD28 clone 1A7 is shown. [Figure 17E] The sequence of the affinity-optimized variable weight domain derived from anti-CD28 clone 1A7 is shown. [Figure 17F] The sequence of the affinity-optimized variable weight domain derived from anti-CD28 clone 1A7 is shown. It should be noted that the variable weight domain can pair with any of the other variable light domains shown in Figures 16 and 18. [Figure 18A] The sequence of the affinity-optimized variable light domain derived from anti-CD28 clone 1A7 is shown. [Figure 18B] The sequence of the affinity-optimized variable light domain derived from anti-CD28 clone 1A7 is shown. [Figure 18C] The sequence of the affinity-optimized variable light domain derived from anti-CD28 clone 1A7 is shown. [Figure 18D] The sequence of the affinity-optimized variable light domain derived from anti-CD28 clone 1A7 is shown. [Figure 18E] The sequence of the affinity-optimized variable light domain derived from anti-CD28 clone 1A7 is shown. [Figure 18F] The sequence of the affinity-optimized variable light domain derived from anti-CD28 clone 1A7 is shown. [Figure 18G] The sequence of the affinity-optimized variable light domain derived from anti-CD28 clone 1A7 is shown. [Figure 18H] The sequence of the affinity-optimized variable light domain derived from anti-CD28 clone 1A7 is shown. [Figure 18I] The sequence of the affinity-optimized variable light domain derived from anti-CD28 clone 1A7 is shown. It should be noted that the variable heavy domain can pair with any of the other variable heavy domains shown in Figures 16 and 17. [Figure 19A] The following shows an exemplary affinity-optimized sequence of 1A7 VH / VL pairs. [Figure 19B] The following shows an exemplary affinity-optimized sequence of 1A7 VH / VL pairs. [Figure 19C] Exemplary affinity-optimized 1A7 VH / VL pair sequences are shown. It should be noted that these pairs can be formalized as Fab or scFv. Furthermore, in scFv form, these pairs can be formalized with VHVL orientation or VLVH orientation. [Figure 20A] The consensus framework region (FR) and complementarity determination region (CDR) (in Kabat) of the variable weight and variable light domain variants of anti-CD28 clone 1A7 are shown. [Figure 20B] The consensus framework region (FR) and complementarity determination region (CDR) (in Kabat) of the variable weight and variable light domain variants of anti-CD28 clone 1A7 are shown. [Figure 21] This shows exemplary affinity operations 1A7, specifically the binding affinity in relation to VH / VL pairs and scFv (in relation to the 1+1 Fab-scFv-FcbsAb form). [Figure 22A]The variable heavy chain sequence and variable light chain sequence of the additional CD28-binding domain used in TROP2×CD28 bsAb of the present invention are shown. [Figure 22B] The variable heavy chain sequence and variable light chain sequence of the additional CD28-binding domain used in TROP2×CD28 bsAb of the present invention are shown. [Figure 22C] The variable heavy chain sequence and variable light chain sequence of the additional CD28-binding domain used in TROP2×CD28 bsAb of the present invention are shown. [Figure 22D] The variable heavy chain sequence and variable light chain sequence of the additional CD28-binding domain used in TROP2×CD28 bsAb of the present invention are shown. [Figure 22E] The variable heavy chain sequence and variable light chain sequence of the additional CD28-binding domain used in TROP2×CD28 bsAb of the present invention are shown. [Figure 22F] The variable heavy chain sequence and variable light chain sequence of the additional CD28-binding domain used in TROP2×CD28 bsAb of the present invention are shown. [Figure 22G] The variable heavy chain sequence and variable light chain sequence of the additional CD28-binding domain used in TROP2×CD28 bsAb of the present invention are shown. [Figure 22H] The variable heavy chain and variable light chain sequences of the additional CD28-binding domain used in TROP2×CD28 bsAb of the present invention are shown. As is the case with all sequences described herein that contain CDRs, the precise location of the CDR may vary slightly depending on the numbering used, as shown in Table 2. Therefore, this specification includes not only the underlined CDRs but also CDRs contained within the VH and VL domains using other numbering systems. Furthermore, for all sequences in the figures, these VH and VL sequences may be used in either scFv or Fab format. [Figure 23A] The following shows exemplary variable heavy chain sequences and variable light chain sequences of the TROP2-binding domain used in the TROP2×CD28 bsAb of the present invention. [Figure 23B]The following shows exemplary variable heavy chain sequences and variable light chain sequences of the TROP2-binding domain used in the TROP2×CD28 bsAb of the present invention. [Figure 23C] The following shows exemplary variable heavy chain sequences and variable light chain sequences of the TROP2-binding domain used in the TROP2×CD28 bsAb of the present invention. [Figure 23D] The following shows exemplary variable heavy chain sequences and variable light chain sequences of the TROP2-binding domain used in the TROP2×CD28 bsAb of the present invention. [Figure 23E] The following shows exemplary variable heavy chain sequences and variable light chain sequences of the TROP2-binding domain used in the TROP2×CD28 bsAb of the present invention. [Figure 23F] The following shows exemplary variable heavy chain sequences and variable light chain sequences of the TROP2-binding domain used in the TROP2×CD28 bsAb of the present invention. [Figure 23G] The following shows exemplary variable heavy chain sequences and variable light chain sequences of the TROP2-binding domain used in the TROP2×CD28 bsAb of the present invention. [Figure 23H] The following shows exemplary variable heavy chain sequences and variable light chain sequences of the TROP2-binding domain used in the TROP2×CD28 bsAb of the present invention. [Figure 23I] The following shows exemplary variable heavy chain sequences and variable light chain sequences of the TROP2-binding domain used in the TROP2×CD28 bsAb of the present invention. [Figure 23J] The following shows exemplary variable heavy chain sequences and variable light chain sequences of the TROP2-binding domain used in the TROP2×CD28 bsAb of the present invention. [Figure 23K] The following shows exemplary variable heavy chain sequences and variable light chain sequences of the TROP2-binding domain used in the TROP2×CD28 bsAb of the present invention. [Figure 23L] The following shows exemplary variable heavy chain sequences and variable light chain sequences of the TROP2-binding domain used in the TROP2×CD28 bsAb of the present invention. [Figure 23M]The variable heavy chain and variable light chain sequences of the TROP2 binding domain used in TROP2×CD28 bsAb of the present invention are shown. As is the case with all sequences described herein that contain a CDR, the location of the CDR may be determined according to the numbering scheme shown in Table 2. Furthermore, for all sequences in the figure, these VH and VL sequences may be used in either scFv or Fab format. [Figure 24] The variable heavy chain and variable light chain sequences of exemplary TROP2-binding domains used in TROP2×CD28 bsAb of the present invention are shown. Each of the variable heavy domains in this figure can be paired with any of the variable light domains in this figure. As is the case with all sequences described herein that contain a CDR, the location of the CDR can be determined according to the numbering scheme shown in Table 2. Furthermore, for all sequences in the figure, these VH and VL sequences can be used in either scFv or Fab format. [Figure 25] The variable heavy chain and variable light chain sequences of exemplary TROP2-binding domains used in TROP2×CD28 bsAb of the present invention are shown. Each of the variable heavy domains in this figure can be paired with any of the variable light domains in this figure. As is the case with all sequences described herein that contain a CDR, the location of the CDR can be determined according to the numbering scheme shown in Table 2. Furthermore, for all sequences in the figure, these VH and VL sequences can be used in either scFv or Fab format. [Figure 26] The variable heavy chain and variable light chain sequences of exemplary TROP2-binding domains used in TROP2×CD28 bsAb of the present invention are shown. Each of the variable heavy domains in this figure can be paired with any of the variable light domains in this figure. As is the case with all sequences described herein that contain a CDR, the location of the CDR can be determined according to the numbering scheme shown in Table 2. Furthermore, for all sequences in the figure, these VH and VL sequences can be used in either scFv or Fab format. [Figure 27]The variable heavy chain and variable light chain sequences of exemplary TROP2-binding domains used in TROP2×CD28 bsAb of the present invention are shown. Each of the variable heavy domains in this figure can be paired with any of the variable light domains in this figure. As is the case with all sequences described herein that contain a CDR, the location of the CDR can be determined according to the numbering scheme shown in Table 2. Furthermore, for all sequences in the figure, these VH and VL sequences can be used in either scFv or Fab format. [Figure 28A] The following shows exemplary variable heavy chain sequences and variable light chain sequences of the TROP2-binding domain used in the TROP2×CD28 bsAb of the present invention. [Figure 28B] The variable heavy chain and variable light chain sequences of exemplary TROP2-binding domains used in TROP2×CD28 bsAb of the present invention are shown. Each of the variable heavy domains in this figure can be paired with any of the variable light domains in this figure. As is the case with all sequences described herein that contain a CDR, the location of the CDR can be determined according to the numbering scheme shown in Table 2. Furthermore, for all sequences in the figure, these VH and VL sequences can be used in either scFv or Fab format. [Figure 29]Figures A and B demonstrate, A) classical T cell / APC interaction, and B) reproduction of classical T cell / APC interaction by combining CD3 bispecific antibodies and CD28 bispecific antibodies. In classical T cell / APC interaction, there is a first signal (signal 1) provided by TCR reactivity with peptide-MHC and a second signal (signal 2) provided by crosslinking of CD28 by CD80 / CD86 expressed on APCs, which together fully activate T cells. In contrast, treatment with CD3 bispecificity provides only the first signal. The CD28 signal may be provided by CD28 bispecificity with the intention of promoting activation and proliferation via CD28 costimulation. In some embodiments, TAA1 and TAA2 may be different antigens. In some embodiments, TAA1 and TAA2 may be the same antigen but may have different epitopes. In some embodiments, TAA1 and TAA2 may be the same antigen and have the same epitope. [Figure 30A] The sequence of TROP2×CD28 bsAb, an example of the 1+1 Fab-scFv-Fc format, is shown. [Figure 30B] The sequence of an exemplary TROP2×CD28 bsAb in the 1+1 Fab-scFv-Fc format is shown. Slashes indicate the boundary(s) between the variable domain and other domains, e.g., the domain linker and the constant domain. In addition, the nomenclature convention indicates the orientation of the scFv from the N-terminus to the C-terminus. As is the case with all sequences described herein that contain a CDR, the location of the CDR may be determined according to the numbering scheme shown in Table 2. It should be noted that TROP2×CD28 bsAb can utilize variable, Fc, and constant domain sequences that are 90, 95, 98, and 99% identical (as defined herein), and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. In addition, each sequence outlined herein may or may not contain the M428L / N434S variant in one or preferably both Fc domains, and the inclusion of M428L / N434S results in a longer half-life in serum. [Figure 31] The sequence of an exemplary TROP2×CD28 bsAb in 2+1 mAb-scFv format is shown. Slashes indicate the boundary(s) between the variable domain and other domains, e.g., the domain linker and the constant domain. In addition, the nomenclature convention indicates the orientation of the scFv from the N-terminus to the C-terminus. As is the case with all sequences described herein that contain a CDR, the location of the CDR may be determined according to the numbering scheme shown in Table 2. It should be noted that TROP2×CD28 bsAb can utilize variable, Fc, and constant domain sequences that are 90, 95, 98, and 99% identical (as defined herein), and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. In addition, each sequence outlined herein may or may not contain the M428L / N434S variant in one or preferably both Fc domains, and the inclusion of M428L / N434S results in a longer half-life in serum. [Figure 32A] The following is an exemplary sequence of an anti-CD3 binding domain suitable for use in a CD3 bispecific antibody that can be combined with the CD28 bispecific antibody of the present invention. [Figure 32B] The following is an exemplary sequence of an anti-CD3 binding domain suitable for use in a CD3 bispecific antibody that can be combined with the CD28 bispecific antibody of the present invention. [Figure 32C] The following is an exemplary sequence of an anti-CD3 binding domain suitable for use in a CD3 bispecific antibody that can be combined with the CD28 bispecific antibody of the present invention. [Figure 32D] The following is an exemplary sequence of an anti-CD3 binding domain suitable for use in a CD3 bispecific antibody that can be combined with the CD28 bispecific antibody of the present invention. [Figure 32E] The following is an exemplary sequence of an anti-CD3 binding domain suitable for use in a CD3 bispecific antibody that can be combined with the CD28 bispecific antibody of the present invention. [Figure 32F]The following is a sequence of an exemplary anti-CD3 binding domain suitable for use in a CD3 bispecific antibody that can be combined with the CD28 bispecific antibody of the present invention. The CDR is indicated by an underline, the scFv linker by a double underline (in the sequence, the scFv linker is the positively charged scFv(GKPGS)4 linker (SEQ ID NO: 24), but as will be understood to those skilled in the art, this linker can be replaced by other linkers, including uncharged or negatively charged linkers, some of which are shown in Figure 6), and a slash indicates the boundary(s) of a variable domain. In addition, the nomenclature convention indicates the orientation of the scFv from the N-terminus to the C-terminus. As is true for all sequences described herein that contain a CDR, the precise location of the CDR may vary slightly depending on the numbering used, as shown in Table 2; therefore, this specification includes not only underlined CDRs but also CDRs contained within the VH and VL domains using other numbering systems. Furthermore, for all sequences shown in the figure, these VH and VL sequences can be used in either scFv or Fab format. [Figure 33] A and B show the induction of IL2 secretion from purified T cells incubated with αCD3-transfected A431-β21-null cells in an effector:target ratio of A) 10:1 or B) 1:1. [Figure 34A] This shows a novel 1A7 VH sequence that has been affinity-modified to pair with the IGKV1-39 germline sequence, as well as its consensus framework region (FR) and complementarity-determining region (CDR). [Figure 34B] This shows a novel 1A7 VH sequence that has been affinity-modified to pair with the IGKV1-39 germline sequence, as well as its consensus framework region (FR) and complementarity-determining region (CDR). [Figure 34C] This shows a novel 1A7 VH sequence that has been affinity-modified to pair with the IGKV1-39 germline sequence, as well as its consensus framework region (FR) and complementarity-determining region (CDR). [Figure 35] This study demonstrates the affinity of a novel 1A7 variant paired with IGKV1-39 human germline VL to human CD28. [Figure 36A] This shows a novel TROP2-binding domain clone, 1F11-1A3.315, and an exemplary affinity-modified VH variant that can pair with L1 (it should be noted that L1 is the IGKV1-39 germline sequence). [Figure 36B] This shows a novel TROP2-binding domain clone, 1F11-1A3.315, and an exemplary affinity-modified VH variant that can pair with L1 (it should be noted that L1 is the IGKV1-39 germline sequence). [Figure 36C] The novel TROP2-binding domain clone 1F11-1A3.315 and an exemplary affinity-manipulated VH variant that can pair with L1 (it should be noted that L1 is the IGKV1-39 germline sequence) are shown. Furthermore, the consensus framework region (FR) and complementarity-determining region (CDR) of the 1F11-1A3.315 VH are also shown. As is the case with all sequences described herein that contain a CDR, the precise location of the CDR may vary slightly depending on the numbering used, as shown in Table 2; therefore, this specification includes not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems. Furthermore, for all sequences in the figures, these VH and VL sequences may be used in either scFv or Fab format. [Figure 37] The KD binding constant, association constant (ka), and dissociation constant (kd) of affinity-modified anti-TROP2 clone 1F11-1A3.315 to human and cynomolgus monkey TROP2 are shown. [Figure 38A] The sequence of an exemplary TROP2×CD28 bsAb in the 1+1 CLC form is shown. [Figure 38B] The sequence of an exemplary TROP2×CD28 bsAb in the 1+1 CLC form is shown. [Figure 38C] The sequence of an exemplary TROP2×CD28 bsAb in the 1+1 CLC form is shown. [Figure 38D] The sequence of an exemplary TROP2×CD28 bsAb in the 1+1 CLC form is shown. [Figure 38E] An exemplary sequence of TROP2×CD28 bsAb in 1+1 CLC form is shown. The CDR is underlined, and slashes indicate the boundary(s) between the variable region, linker, Fc region, and constant domain. It should be noted that TROP2×CD28 bsAb can utilize variable region, Fc region, and constant domain sequences that are 90, 95, 98, and 99% identical (as defined herein), and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. In addition, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, thereby increasing the half-life in serum. [Figure 39A] The sequence of a typical TROP2×CD28 bsAb in the 2+1 CLC form is shown. [Figure 39B] The sequence of a typical TROP2×CD28 bsAb in the 2+1 CLC form is shown. [Figure 39C] The sequence of a typical TROP2×CD28 bsAb in the 2+1 CLC form is shown. [Figure 39D] An exemplary sequence of TROP2×CD28 bsAb in 2+1 CLC form is shown. The CDR is underlined, and slashes indicate the boundary(s) between the variable region, linker, Fc region, and constant domain. It should be noted that TROP2×CD28 bsAb can utilize variable region, Fc region, and constant domain sequences that are 90, 95, 98, and 99% identical (as defined herein), and / or contain 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid substitutions. In addition, each sequence outlined herein may contain or exclude the M428L / N434S variant in one or preferably both Fc domains, thereby increasing the half-life in serum. [Figure 40A] The sequence of the RSV control (i.e., bsAb in which the TROP2 or CD28 binding domain is replaced with a "silent" RSV binding domain) is shown. [Figure 40B]The sequences of RSV controls (i.e., bsAb in which the TROP2 or CD28 binding domain is replaced with a “silent” RSV binding domain) are shown. These controls are used for a variety of experimental purposes, including serving as a surrogate for how they behave on TROP2-negative cells. [Figure 41] This shows the induction of IL-2 secretion when 1 μg / mL of exemplary B7H3×CD3 bsAb is combined with various TROP2×CD28 bsAbs with different TROP2 and CD28 binding affinities in the presence of tumor cells (A)647-V (high TROP), B)A431 (medium TROP), C)MDA-MB-231 (low-medium TROP), and D)DU145 (low-medium TROP) (1:10 effector:target). [Figure 42] This study demonstrates the induction of IL-2 secretion when an exemplary B7H3×CD3 bsAb at 1 μg / mL is combined with various forms of TROP2×CD28 bsAb having different TROP2 binding affinities, in the presence of OVCAR5 tumor cells (1:1 effector:target). [Figure 43] This study demonstrates the induction of IL-2 secretion when an exemplary B7H3×CD3 bsAb at 1 μg / mL is combined with various forms of TROP2×CD28 bsAbs with different TROP2 and CD28 binding affinities, in the presence of OVCAR5 tumor cells (1:1 effector:target). [Figure 44] This shows the induction of IL-2 secretion when 1 μg / mL of exemplary B7H3×CD3 bsAb is combined with various TROP2×CD28 bsAbs with different TROP2 and CD28 binding affinities in the presence of A)DU145, B)MCF7, C)OVCAR5, and D)MDA-MB-231 tumor cells (1:1 effector:target). [Figure 45] The EC50 of induction of IL-2 secretion when 1 μg / mL of exemplary B7H3×CD3 bsAb is combined with various TROP2×CD28 bsAbs with different TROP2 and CD28 binding affinities in the presence of DU145, MDA-MB-231, OVCAR5, or MCF7 tumor cells (1:1 effector:target) is shown. [Figure 46]This study demonstrates the induction of IL-2 secretion in the presence of MDA-MB-231 tumor cells (1:10 effector:target) when TROP2×CD28 bsAb is administered alone or in combination with 10 μg / ml PD-1 mAb. The data show that the activity of TROP×CD28 bsAb is enhanced when combined with PD-1 mAb. [Figure 47] This shows the induction of T cell activation (A), T cell proliferation (indicated by ki67 (B) and T cell count (C)) and tumor cell killing (D) when 1 μg / ml of TROP2×CD28 bsAb XENP44599 is incubated with a co-culture of T cells and OVCAR5 cells and with exemplary titration doses of B7H3×CD3 bsAb. [Figure 48] This shows the induction of IFNγ secretion when 1 μg / mL of exemplary B7H3×CD3 bsAb is combined with various TROP2×CD28 bsAbs with different TROP2 and CD28 binding affinities in the presence of A)DU145, B)MCF7, C)OVCAR5, and D)MDA-MB-231 tumor cells (1:1 effector:target). [Figure 49] This shows the induction of IL-2 secretion when exemplary titration doses of B7H3×CD3 bsAb are combined with 1 μg / ml TROP2×CD28 bsAb having various TROP2 and CD28 binding affinities in the presence of tumor cells (A)647-V (high TROP), B)A431 (medium TROP), C)MDA-MB-231 (low-medium TROP), and D)DU145 (low-medium TROP) (1:10 effector:target). [Figure 50] This study demonstrates the induction of IL-2 secretion when an exemplary B7H3×CD3 bsAb at 1 μg / mL is combined with various forms of TROP2×CD28 bsAb having different TROP2 binding affinities, in the presence of pp65-MDA-MB-231 tumor cells (1:1 effector:target). [Figure 51]This study demonstrates the induction of IL-2 secretion when an exemplary B7H3×CD3 bsAb at 1 μg / mL is combined with various forms of TROP2×CD28 bsAbs having different TROP2 and CD28 binding affinities in the presence of pp65-MDA-MB-231 tumor cells (1:1 effector:target). [Figure 52] This shows a novel TROP2-binding domain clone, 2C5A3.316 (it should be noted that L1 is the IGKV1-39 germline sequence). As is the case with all sequences described herein that contain CDRs, the precise location of the CDR may vary slightly depending on the numbering used, as shown in Table 2. Therefore, this specification includes not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems. Furthermore, for all sequences in the figure, these VH and VL sequences may be used in either scFv or Fab format. [Figure 53] This shows a novel TROP2-binding domain clone 1A3A4.312 (it should be noted that L1 is the IGKV1-39 germline sequence). As is the case with all sequences described herein that contain CDRs, the precise location of the CDR may vary slightly depending on the numbering used, as shown in Table 2. Therefore, this specification includes not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems. Furthermore, for all sequences in the figure, these VH and VL sequences may be used in either scFv or Fab format. [Figure 54]This shows a novel TROP2-binding domain clone, 1B2A4.312 (it should be noted that L1 is the IGKV1-39 germline sequence). As is the case with all sequences described herein that contain CDRs, the precise location of the CDR may vary slightly depending on the numbering used, as shown in Table 2. Therefore, this specification includes not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems. Furthermore, for all sequences in the figure, these VH and VL sequences may be used in either scFv or Fab format. [Figure 55] This shows a novel TROP2-binding domain clone, 1B11A3.316 (it should be noted that L1 is the IGKV1-39 germline sequence). As is the case with all sequences described herein that contain CDRs, the precise location of the CDR may vary slightly depending on the numbering used, as shown in Table 2. Therefore, this specification includes not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems. Furthermore, for all sequences in the figure, these VH and VL sequences may be used in either scFv or Fab format. [Figure 56] This shows a novel TROP2-binding domain clone, 1C9A4.313 (it should be noted that L1 is the IGKV1-39 germline sequence). As is the case with all sequences described herein that contain CDRs, the precise location of the CDR may vary slightly depending on the numbering used, as shown in Table 2. Therefore, this specification includes not only underlined CDRs but also CDRs contained within VH and VL domains using other numbering systems. Furthermore, for all sequences in the figure, these VH and VL sequences may be used in either scFv or Fab format. [Figure 57] The serum concentrations of XENP44599 and XENP44595 in cynomolgus monkeys over time are shown. [Figure 58]This study demonstrates the induction of IL-2 secretion by purified T cells incubated with 1 μg / ml exemplary EpCAM×CD3 bsAb in combination with various TROP2×CD28 bsAbs having different TROP2 and CD28 binding affinities, in the presence of OVCAR5 TROP2-high tumor cells (1:1 effector:target ratio). [Figure 59A] This study demonstrates the induction of IL-2 secretion by purified T cells incubated with 1 μg / ml exemplary EpCAM×CD3 bsAb in combination with various TROP2×CD28 bsAb with different TROP2 and CD28 binding affinities, in the presence of OVCAR5 high TROP2 or OVCAR3 low TROP2 tumor cells (1:1 effector:target ratio), along with changes in potency and efficacy in each cell line. [Figure 59B] This study demonstrates the induction of IL-2 secretion by purified T cells incubated with 1 μg / ml exemplary EpCAM×CD3 bsAb in combination with various TROP2×CD28 bsAb with different TROP2 and CD28 binding affinities, in the presence of OVCAR5 high TROP2 or OVCAR3 low TROP2 tumor cells (1:1 effector:target ratio), along with changes in potency and efficacy in each cell line. [Figure 59C] This study demonstrates the induction of IL-2 secretion by purified T cells incubated with 1 μg / ml exemplary EpCAM×CD3 bsAb in combination with various TROP2×CD28 bsAb with different TROP2 and CD28 binding affinities, in the presence of OVCAR5 high TROP2 or OVCAR3 low TROP2 tumor cells (1:1 effector:target ratio), along with changes in potency and efficacy in each cell line. [Figure 59D] This study demonstrates the induction of IL-2 secretion by purified T cells incubated with 1 μg / ml exemplary EpCAM×CD3 bsAb in combination with various TROP2×CD28 bsAb with different TROP2 and CD28 binding affinities, in the presence of OVCAR5 high TROP2 or OVCAR3 low TROP2 tumor cells (1:1 effector:target ratio), along with changes in potency and efficacy in each cell line. [Figure 59E]This study demonstrates the induction of IL-2 secretion by purified T cells incubated with 1 μg / ml exemplary EpCAM×CD3 bsAb in combination with various TROP2×CD28 bsAb with different TROP2 and CD28 binding affinities, in the presence of OVCAR5 high TROP2 or OVCAR3 low TROP2 tumor cells (1:1 effector:target ratio), along with changes in potency and efficacy in each cell line. [Figure 59F] This study demonstrates the induction of IL-2 secretion by purified T cells incubated with 1 μg / ml exemplary EpCAM×CD3 bsAb in combination with various TROP2×CD28 bsAb with different TROP2 and CD28 binding affinities, in the presence of OVCAR5 high TROP2 or OVCAR3 low TROP2 tumor cells (1:1 effector:target ratio), along with changes in potency and efficacy in each cell line. [Figure 59G] This study demonstrates the induction of IL-2 secretion by purified T cells incubated with 1 μg / ml exemplary EpCAM×CD3 bsAb in combination with various TROP2×CD28 bsAb with different TROP2 and CD28 binding affinities, in the presence of OVCAR5 high TROP2 or OVCAR3 low TROP2 tumor cells (1:1 effector:target ratio), along with changes in potency and efficacy in each cell line. [Figure 59H] This study demonstrates the induction of IL-2 secretion by purified T cells incubated with 1 μg / ml exemplary EpCAM×CD3 bsAb in combination with various TROP2×CD28 bsAb with different TROP2 and CD28 binding affinities, in the presence of OVCAR5 high TROP2 or OVCAR3 low TROP2 tumor cells (1:1 effector:target ratio), along with changes in potency and efficacy in each cell line. [Figure 59I] This study demonstrates the induction of IL-2 secretion by purified T cells incubated with 1 μg / ml exemplary EpCAM×CD3 bsAb in combination with various TROP2×CD28 bsAb with different TROP2 and CD28 binding affinities, in the presence of OVCAR5 high TROP2 or OVCAR3 low TROP2 tumor cells (1:1 effector:target ratio), along with changes in potency and efficacy in each cell line. [Figure 60] This shows A) target cell killing and B) T cell proliferation when purified T cells were incubated with exemplary EpCAM×CD3 bsAb at 2.5 ng / mL and a titration dose of TROP2×CD28 bsAb in the presence of OVCAR5 TROP2 hypertumor cells (1:1 E:T ratio; 72 hours). [Modes for carrying out the invention]

[0059] I. Overview TROP2 is a type I transmembrane glycoprotein with signaling function that plays a regulatory role in cell autoogenesis, proliferation, and transformation. TROP2 is overexpressed in several cancers, including endometrial cancer, pancreatic cancer, colorectal cancer, esophageal cancer, cholangiocarcinoma, oral cancer, and glioma. Therefore, TROP2 is a potential target for treating such cancers.

[0060] T cell activation in cancer treatment is a widely studied topic. T cells require multiple signals for complete activation and differentiation. As shown in Figure 29A, signal 1, facilitated by the recognition of peptide-MHC (pMHC) complexes by the T cell receptor (TCR), is absolutely essential for T cell activation. Signal 2, which acts synergistically with and amplifies signal 1, is typically provided by the interaction of CD28 itself with CD80 and CD86, which are ligands for CD28. CD28 engagement alone is typically inactive, but when combined with signal 1 activation, it promotes additional activation, survival, and proliferation signals (including IL-2 secretion). Since CD80 and CD86 are spontaneously expressed only by professional antigen-presenting cells (APCs), the degree of CD28 costimulation in the tumor environment can vary considerably. Therefore, the present invention relates to a novel class of tumor-targeted anti-CD28 × anti-TROP2 bispecific antibodies, wherein the CD80 / CD86 engagement of CD28 mimics the CD80 / CD86 engagement of CD28, thereby providing an artificial source of signal 2 (Figure 29B). Notably, signal 1 may be provided by the innate TCR:pMHC recognition of tumor cells, or by a combination of CD28 bispecificity and CD3 bispecificity that can mimic signal 1 (e.g., anti-CD3 × anti-TROP2).

[0061] Accordingly, this specification provides novel anti-CD28 × anti-TROP2 (also known as "αCD28 × αTROP2" and sometimes "CD28 × TROP2") bispecific antibodies and methods for using such antibodies for the treatment of TROP2-associated cancers. In many cases, these bispecific antibodies are heterodimers. The αCD28 × αTROP2 antibodies of interest can agonist-conjugate to CD28 costimulatory molecules on T cells and target TROP2 on TROP2-expressing tumor cells. Thus, such antibodies selectively enhance antitumor activity at TROP2-expressing tumor sites while minimizing peripheral toxicity. The antibodies of interest provided herein are particularly useful for enhancing antitumor activity, whether used alone as monotherapy or in combination with other anticancer therapies more fully described herein.

[0062] Accordingly, in one embodiment, what is provided herein is a heterodimer antibody that binds to two different antigens, for example, the antibody is "bispecific" in that it binds to two different target antigens, generally CD28 and TROP2, as described below. These heterodimer antibodies can bind to each of the target antigens in either a monovalent (e.g., having a single antigen-binding domain, e.g., a pair of variable heavy and variable light domains) or a bivalent (having two antigen-binding domains, each independently binding to the antigen). In some embodiments, the heterodimer antibody provided herein comprises one CD28-binding domain and one TROP2-binding domain (e.g., a heterodimer antibody of the "1+1 Fab-scFv-Fc" form described herein, and thus bispecific and bivalent). In other embodiments, the heterodimer antibodies provided herein contain one CD28-binding domain and two TROP2-binding domains (e.g., the "2+1 Fab2-scFv-Fc" heterodimer antibody described herein, which is bispecific but trivalent because it contains three antigen-binding domains (ABD)). The heterodimer antibodies provided herein are based on the use of different monomers containing amino acid substitutions (i.e., "scuba riants") that "bias" towards heterodimer formation rather than homodimer formation, as will be more thoroughly outlined below. In some embodiments, the heterodimer antibodies are also combined with a purification variant (e.g., a "pI variant") that allows for the simple purification of heterodimers from homodimers, as will also be outlined below. The heterodimer bispecific antibodies provided typically rely on the use of engineered or variant Fc domains that can be self-constructed in generative cells to produce heterodimer proteins, and on methods for producing and purifying such heterodimer proteins.

[0063] II. Nomenclature The nomenclature for specific antigen-binding domains (e.g., TROP2 and CD28-binding domains) uses the format "Hx.xx_Ly.yy", where the numbers are unique identifiers for specific variable-chain sequences. For example, the CD28-binding domain "1A7[CD28]_H1_L1" (Figure 16) contains a variable heavy domain, H1, and a variable light domain, L1. When these sequences are used as scFv, the name "H1_L1" indicates that the binding domain contains a combination of the variable heavy domain "H1" and the variable light domain "L1," with an N-terminal to C-terminal orientation of VH-linker-VL. A molecule having the same sequences of heavy and light variable domains but in the reverse order (N-terminal to C-terminal VL-linker-VH orientation) would be designated "L1_H1". Similarly, different constructs can "mix and match" heavy and light chains, as is evident from the sequence listings and figures.

[0064] III. Definition To ensure that this application can be fully understood, some definitions are provided below. Such definitions are intended to encompass grammatical equivalents.

[0065] In this specification, “CD28,” “differentiation antigen group 28,” and “Tp44” (e.g., Genebank accession numbers NP_001230006 (human), NP_001230007 (human), NP_006130 (human), and NP_031668 (mouse)) refer to B7 receptors expressed on T cells that provide costimulatory signals necessary for T cell activation and survival. T cell stimulation via CD28, in addition to the T cell receptor (TCR), provides a potent signal for the production of various interleukins. CD28 is a receptor for the CD80 (B7.1) and CD86 (B7.2) proteins. CD28 contains an intracellular domain with an SH2 domain-containing protein, in particular, a YMNM motif (SEQ ID NO: 749) that is important for the recruitment of PI3K. CD28 also contains two proline-rich motifs to which SH3-containing proteins can be bound. An exemplary CD28 sequence is shown in Figure 1. Unless otherwise stated, references to CD28 refer to human CD28 sequences.

[0066] In this specification, “TROP2,” “tumor-associated calcium signaling transducer 2,” “TACSTD2,” “epidermal glycoprotein 1,” and “EGP-1” refer to transmembrane glycoproteins that are part of the tumor-associated calcium signaling transducer (TACSTD) family. TROP2 has signaling function and plays a regulatory role in cell autoogenesis, proliferation, and transformation. The TROP2 sequence is shown, for example, in Figure 2. TROP2 is expressed in certain cancers, including endometrial cancer, pancreatic cancer, colorectal cancer, esophageal cancer, cholangiocarcinoma, oral cancer, and glioma.

[0067] In this specification, "reduction" means a decrease or removal of activity. Therefore, for example, "reduction of FcγR binding" means that the Fc region amino acid variant has less than 50% of the initial binding compared to an Fc region that does not contain the particular variant, with a loss of activity of more than 70-80-90-95-98% being preferred, and is usually below the binding level detectable in Biacore, SPR, or BLI assays. Those particularly used in FcγR binding reduction are shown in Figure 5, and these are usually attached to both monomers.

[0068] "ADCC," or antibody-dependent cell-mediated cytotoxicity, as used herein, refers to a cell-mediated reaction in which nonspecific cytotoxic cells expressing FcγR recognize a bound antibody on a target cell, subsequently causing lysis of the target cell. ADCC correlates with binding to FcγRIIIa; increased binding to FcγRIIIa results in increased ADCC activity.

[0069] When used herein, "ADCP" or "antibody-dependent cell-mediated phagocytosis" refers to a cell-mediated response in which nonspecific phagocytic cells expressing FcγR recognize bound antibodies on target cells, and subsequently cause phagocytosis of the target cells.

[0070] As used herein, the term “antibody” is used in a general sense. The antibodies provided herein may take many forms as described herein, including conventional antibodies and antibody derivatives, fragments, and mimetic compounds described herein.

[0071] Conventional immunoglobulin (Ig) antibodies are "Y"-shaped tetramers. Each tetramer typically consists of two identical polypeptide chain pairs, each pair containing one "light chain" monomer (typically with a molecular weight of about 25 kDa) and one "heavy chain" monomer (typically with a molecular weight of about 50–70 kDa).

[0072] Other useful antibody formats include, but are not limited to, the “1+1 Fab-scFv-Fc” and “2+1 Fab2-scFv-Fc” formats provided herein (see, for example, Figure 15). Additional useful antibody formats include, but are not limited to, the “1+1 common light chain” and “2+1 common light chain” “mAb-Fv” “mAb-scFv” “Central-Fv” “1-arm scFv-mAb” “scFv-mAb” “Dual scFv” and “Trident” format antibodies (Figure 15). See also US20180127501A1 (which is incorporated herein by reference, particularly the relevant parts concerning antibody formats) (see, for example, Figure 2 of US20180127501A1).

[0073] Antibody heavy chains typically include a variable weight (VH) domain containing vhCDR1-3 and an Fc domain containing CH2-CH3 monomers. In some embodiments, antibody heavy chains also include hinge and CH1 domains. Conventional antibody heavy chains are monomers organized from N to C-terminus: VH-CH1-hinge-CH2-CH3. The CH1-hinge-CH2-CH3 are collectively referred to as the "constant domain" or "constant region" of the heavy chain of antibodies in which five different categories or "isotypes" exist: IgA, IgD, IgG, IgE, and IgM.

[0074] In some embodiments, the antibodies provided herein include a constant IgG isotype domain, which has several subclasses, including but not limited to IgG1, IgG2, IgG3, and IgG4. In the IgG subclasses of immunoglobulins, several immunoglobulin domains are present in the heavy chain. As used herein, “immunoglobulin (Ig) domain” means a region of immunoglobulin having a distinctive tertiary structure. Heavy chain domains, including constant heavy (CH) domains and hinge domains, are of interest in the present invention. In relation to IgG antibodies, each IgG isotype has three CH domains. Thus, in relation to IgG, the “CH” domains are as follows: “CH1” points to positions 118-215 according to the EU index in Kabat; “Hinge” points to positions 216-230 according to the EU index in Kabat; “CH2” points to positions 231-340 according to the EU index in Kabat; and “CH3” points to positions 341-447 according to the EU index in Kabat. As shown in Figure 1, the exact numbering and arrangement of heavy chain domains may differ between different numbering systems. As shown herein and described below, pI variants may be in one or more CH regions and in the hinge regions discussed below.

[0075] It should be noted that IgG1 has different allotypes with polymorphisms at 356(D or E) and 358(L or M). The sequences shown herein use the 356E / 358M allotype, but other allotypes are included herein. That is, any sequence containing the IgG1 Fc domain included herein may have 356D / 358L, which replaces the 356E / 358M allotype. It should also be understood that therapeutic antibodies may also contain isotype and / or subclass hybrids. For example, as shown in U.S. Publication 2009 / 0163699 (incorporated by reference), the antibody in some embodiments contains a human IgG1 / G2 hybrid.

[0076] When used herein, “Fc,” “Fc region,” or “Fc domain” refers to a polypeptide containing the constant region of an antibody, in some examples all or part of the first constant region immunoglobulin domain (e.g., CH1), and in some cases optionally all or part of the hinge. In the case of IgG, the Fc domain includes the immunoglobulin domains CH2 and CH3 (Cγ2 and Cγ3), and optionally all or part of the hinge region between CH1 (Cγ1) and CH2 (Cγ2). Thus, in some cases, the Fc domain includes CH2-CH3 and hinge-CH2-CH3 from N to C-terminus. In some embodiments, the Fc domain is from IgG1, IgG2, IgG3, or IgG4, with IgG1 hinge-CH2-CH3 and IgG4 hinge-CH2-CH3 being particularly utilized in many embodiments. Also, in the case of the human IgG1 Fc domain, the hinge may include the C220S amino acid substitution. Furthermore, in the case of the human IgG4 Fc domain, the hinge may include the S228P amino acid substitution. While the boundaries of the Fc region can vary, the human IgG heavy chain Fc region is typically defined as containing residues E216, C226, or A231 at its carboxyl terminus, and the numbering follows the EU index in Kabat. In some embodiments, amino acid modifications are made to the Fc region to alter binding to one or more FcγR or FcRn, as more fully described below.

[0077] In this specification, “heavy chain constant region” means the CH1-hinge-CH2-CH3 portion of an antibody (or fragment thereof), excluding the variable heavy domain; in the EU numbering of human IgG1, this corresponds to amino acids 118-447. In this specification, “heavy chain constant region fragment” means a heavy chain constant region containing fewer amino acids from either or both of the N and C-terminuses, but still retaining the ability to form dimers with another heavy chain constant region.

[0078] Another type of heavy chain domain is the hinge region. In this specification, “hinge,” “hinge region,” “antibody hinge region,” or “hinge domain” refers to a flexible polypeptide containing amino acids between the first and second constant domains of an antibody. Structurally, the IgG CH1 domain terminates at EU position 215, and the IgG CH2 domain begins at residue EU position 231. Thus, in the case of IgG, the antibody hinge is defined herein as containing positions 216 (E216 in IgG1) to 230 (P230 in IgG1), with numbering following the EU index in Kabat. In some cases, “hinge fragments” containing fewer amino acids at either or both of the N and C-terminuses of the hinge domain are used. As described herein, pI variants can also be made from hinge regions. Many of the antibodies herein have at least one cysteine ​​at position 220, following the EU numbering (hinge region), replaced with serine. Typically, this modification is on the “scFv monomer” side (when the 1+1 or 2+1 form is used) for most of the sequences shown herein, but it can also be on the “Fab monomer” side, or both, to reduce disulfide formation. One or both of these replaced cysteines (C220S) are specifically included in the sequences shown herein.

[0079] As will be understood by those skilled in the art, the precise numbering and arrangement of heavy chain constant region domains (i.e., CH1, hinge, CH2, and CH3 domains) may differ between different numbering systems. A useful comparison of heavy constant region numbering according to EU and Kabat is provided below; please refer to Edelman et al., 1969, Proc Natl Acad Sci USA 63:78-85 and Kabat et al., 1991, Sequences of Proteins of Immunological Interest, 5th Ed., United States Public Health Service, National Institutes of Health, Bethesda (incorporated by reference throughout).

[0080] [Table 1]

[0081] Antibody light chains typically contain two domains: a variable light domain (VL) containing light chain CDR vlCDR1-3, and a constant light chain region (often referred to as CL or Cκ). Antibody light chains are typically sequenced VL-CL from the N-terminus to the C-terminus.

[0082] In this specification, “antigen-binding domain” or “ABD” means a set of six complementarity-determining regions (CDRs) that, when present as part of a polypeptide sequence, specifically bind to the target antigens discussed herein (e.g., TROP2 or CD28). As is known in the art, these CDRs generally exist as a first set of variable weight CDRs (vhCDR or VHCDR) and a second set of variable light CDRs (vlCDR or VLCDR), each containing three CDRs: vhCDR1, vhCDR2, vhCDR3 variable weight CDRs and vlCDR1, vlCDR2, vlCDR3, and vhCDR3 variable light CDRs. The CDRs are present in variable weight domains (vhCDR1-3) and variable light domains (vlCDR1-3). The variable weight domains and variable light domains form the Fv region.

[0083] The present invention provides a number of different CDR sets. In this case, a “complete CDR set” includes three variable light CDRs and three variable heavy CDRs, for example, vlCDR1, vlCDR2, vlCDR3, vhCDR1, vhCDR2, and vhCDR3. These may each be part of a larger variable light or variable heavy domain. Furthermore, as will be more fully outlined herein, the variable heavy and variable light domains may be on separate polypeptide chains when heavy and light chains are used (e.g., when Fab is used), or on a single polypeptide chain in the case of scFv sequences.

[0084] As will be understood by those skilled in the art, the exact numbering and arrangement of CDRs may differ between different numbering systems. However, it should be understood that the disclosure of variable weight and / or variable light sequences includes the disclosure of the associated (intrinsic) CDRs. Thus, the disclosure of each variable weight region is a disclosure of vhCDRs (e.g., vhCDR1, vhCDR2, and vhCDR3), and the disclosure of each variable light region is a disclosure of vlCDRs (e.g., vlCDR1, vlCDR2, and vlCDR3). A useful comparison of CDR numbering is as follows: See Lafranc et al., Dev.Comp.Immunol.27(1):55-77(2003):

[0085] [Table 2]

[0086] Throughout this specification, the Kabat numbering system is generally used when referring to residues in the variable domain (approximately residues 1-107 in the light chain variable region and residues 1-113 in the heavy chain variable region), while the EU numbering system is used for the Fc region (e.g., Kabat et al. (cited above) (1991)).

[0087] CDRs contribute to the antigen-binding domain and antigen binding of antibodies, or more specifically, the formation of epitope-binding sites. An "epitope" refers to a determinant that interacts with a specific antigen-binding site in the variable region of an antibody molecule, also known as a paratope. Epitopes are groups of molecules, such as amino acids or sugar side chains, and typically possess specific structural and electrical characteristics. A single antigen may have multiple epitopes.

[0088] An epitope may include amino acid residues directly involved in binding (also known as the immunodominant component of the epitope) and other amino acid residues not directly involved in binding, such as amino acid residues that are effectively blocked by a specific antigen-binding peptide (i.e., those amino acid residues are within the footprint of the specific antigen-binding peptide).

[0089] Epitopes can be either conformational or linear. Conformational epitopes are generated by spatially juxtaposed amino acids from different segments of a linear polypeptide chain. Linear epitopes are generated by adjacent amino acid residues in a polypeptide chain. Conformational and non-conformational epitopes can be distinguished in that binding to the former, rather than the latter, is lost in the presence of a denaturing solvent.

[0090] An epitope typically contains at least three, more commonly five or eight to ten, amino acids in its own spatial conformation. Antibodies that recognize the same epitope can be identified in a simple immunoassay, e.g., "binning," in which one antibody demonstrates its ability to block the binding of another antibody to a target antigen. As outlined below, the present invention includes not only the antigen-binding domains and antibodies listed herein, but also those that compete for binding to the epitopes bound by the listed antigen-binding domains.

[0091] In some embodiments, the six CDRs of the antigen-binding domain are provided by variable heavy and variable light domains. In the "Fab" form, the set of six CDRs consists of two distinct polypeptide sequences: a variable heavy domain (vh or VH; containing vhCDR1, vhCDR2, and vhCDR3) and a variable light domain (vl or VL; containing vlCDR1, vlCDR2, and vlCDR3), where the C-terminus of the vh domain is bound to the N-terminus of the CH1 domain of the heavy chain, and the C-terminus of the vl domain is bound to the N-terminus of the constant light domain (thus forming the light chain). In the scFv form, the vh and vl domains are covalently linked to a single polypeptide sequence, typically via the use of a linker outlined herein ("scFv linker"), which can be either vh-linker-vl or vl-linker-vh (starting from the N-terminus), with the former being the preferred configuration (each side may include an arbitrary domain linker, depending on the form used). Typically, the C-terminus of the scFv domain is linked to the N-terminus of all or part of the hinge in the second monomer.

[0092] When used herein, “variable region” or “variable domain” means a region of immunoglobulin that includes one or more Ig domains substantially encoded by any of the Vκ, Vλ, and / or VH genes constituting kappa, lambda, and each heavy chain immunoglobulin locus, and contains a CDR that confers antigen specificity. Thus, a “variable heavy domain” pairs with a “variable light domain” to form an antigen-binding domain (“ABD”). Each variable domain also includes three hypervariable regions (“complementarity-determining regions,” “CDRs”) (vhCDR1, vhCDR2, and vhCDR3 in the case of a variable heavy domain and vlCDR1, vlCDR2, and vlCDR3 in the case of a variable light domain) and four framework (FR) regions, arranged from the amino terminus to the carboxyl terminus in the following order: FR1-CDR1-FR2-CDR2-FR3-CDR3-FR4.

[0093] "Fab" or "Fab region," as used herein, typically refers to an antibody region on two different polypeptide chains containing VH, CH1, VL, and CL immunoglobulin domains (e.g., VH-CH1 on one chain and VL-CL on the other). Fab may refer to this isolated region or to this region in relation to the bispecific antibody of the present invention. In relation to Fab, Fab includes the Fv region in addition to the CH1 and CL domains.

[0094] When used herein, “Fv,” “Fv fragment,” or “Fv region” means an antibody region containing the VL and VH domains. The Fv region can be formalized as both Fab (two distinct polypeptides, including the constant region typically outlined above, as discussed above) and single-chain Fv (scFv) (where the vl and vh domains are contained in a single peptide conjugated to a linker, typically as discussed herein).

[0095] In this specification, “single-stranded Fv” or “scFv” means a variable weight domain covalently linked to a variable light domain using an scFv linker, typically as discussed herein, to form an scFv or scFv domain. The scFv domain can be oriented either N-to-C-terminus (vh-linker-vl or vl-linker-vh). In sequences shown in the sequence listings and figures, the order of the vh and vl domains is included in the name. For example, H.X_L.Y means that the N-to-C-terminus is vh-linker-vl, and L.Y_H.X means vl-linker-vh.

[0096] Some embodiments of the target antibodies provided herein include at least one scFv domain, which typically contains a variable heavy domain and a variable light domain that are not naturally present but are linked together by an scFv linker. As outlined herein, scFv domains are typically oriented from N to C-terminus as VH-scFv linker-VL, whereas this can be reversed to VL-scFv linker-VH for any scFv domain (or one constructed using vh and vl sequences from Fab) using any linker at one or both ends, depending on the form.

[0097] In this specification, “modification” or “variant” means an amino acid substitution, insertion, and / or deletion in a polypeptide sequence or an alteration to a site chemically linked to a protein. For example, a modification may be an altered carbohydrate or PEG structure linked to a protein. In this specification, “amino acid modification” means an amino acid substitution, insertion, and / or deletion in a polypeptide sequence. For clarity, unless otherwise stated, amino acid modification always refers to amino acids encoded by DNA, e.g., the 20 amino acids that have codons in DNA and RNA.

[0098] In this specification, “amino acid substitution” or “substitution” means the replacement of an amino acid at a specific position in the parent polypeptide sequence with a different amino acid. In particular, in some embodiments, a substitution is a substitution of an amino acid at a specific position that is not found in nature and does not exist naturally in any organism, including within a living organism. For example, substitution E272Y refers to a variant polypeptide, in this case the Fc variant, in which glutamic acid at position 272 is replaced with tyrosine. For clarity, a protein that is manipulated to alter the nucleic acid coding sequence but not alter the starting amino acid (e.g., replacing CGG (encoding arginine) with CGA (still encoding arginine) to increase the host organism’s expression level) is not an “amino acid substitution”; that is, even though it is the generation of a new gene encoding the same protein, if the protein has the same amino acid at the specific position where it starts, it is not an amino acid substitution.

[0099] As used herein, "amino acid insertion" or "insertion" means the addition of an amino acid sequence at a specific position in the parent polypeptide sequence. For example, -233E or 233E specifies the insertion of glutamic acid after position 233 and before position 234. Also, -233ADE or A233ADE specifies the insertion of AlaAspGlu after position 233 and before position 234.

[0100] "Amino acid deletion" or "deletion," as used herein, means the removal of an amino acid sequence at a specific position in the parent polypeptide sequence. For example, E233- or E233#, E233() or E233del specify the deletion of glutamic acid at position 233. Also, EDA233- or EDA233# specifies the deletion of the sequence GluAspAla that begins at position 233.

[0101] "Variant protein," "protein variant," or "variant" as used herein means a protein that differs from that of the parent protein by at least one amino acid modification. A protein variant has at least one amino acid modification compared to the parent protein, but not so many that the variant protein cannot be aligned with the parent protein using an alignment program such as those described below. Typically, a variant protein (e.g., a variant Fc domain outlined herein) is at least 75, 80, 85, 90, 91, 92, 93, 94, 95, 96, 97, 98, or 99% identical to the parent protein using an alignment program such as BLAST, as described below.

[0102] When used herein, "variant" also refers to a specific amino acid modification that confers a particular function (e.g., "heterodimerated variant," "pI variant," "decay variant," etc.).

[0103] As described below, in some embodiments, the parent polypeptide, e.g., the Fc parent polypeptide, is a multiple constant domain or Fc region from a human wild-type sequence, e.g., IgG1, IgG2, IgG3, or IgG4. However, a human sequence having a variant may also function as a “parent polypeptide,” e.g., the IgG1 / 2 hybrid described in U.S. Publication 2006 / 0134105. The protein variant sequences herein preferably have at least about 80% identity with the parent protein sequence, most preferably at least about 90% identity, and more preferably at least about 95-98-99% identity. Therefore, as used herein, “antibody variant” or “variant antibody” means an antibody that differs from the parent antibody by at least one amino acid modification; “IgG variant” or “variant IgG” means an antibody that differs from the parent IgG (which, in this case, is also often derived from a human IgG sequence) by at least one amino acid modification; and “immunoglobulin variant” or “variant immunoglobulin” means an immunoglobulin sequence that differs from the parent immunoglobulin sequence by at least one amino acid modification. “Fc variant” or “variant Fc” means a protein that contains amino acid modifications in its Fc domain compared to the Fc domain of human IgG1, IgG2, or IgG4.

[0104] "Fc variant" or "variant Fc," as used herein, means a protein containing amino acid modifications in the Fc domain. Modifications may be additions, deletions, or substitutions. Fc variants are defined according to the amino acid modifications that constitute them. For example, N434S or 434S is an Fc variant having a serine substitution at position 434 compared to the parent Fc polypeptide, and the numbering follows the EU index. Similarly, M428L / N434S is an Fc variant having the substitutions M428L and N434S compared to the parent Fc polypeptide. WT amino acid identity may not be specified, in which case the aforementioned variant is referred to as 428L / 434S. It should be noted that the order in which substitutions are provided is arbitrary (i.e., for example, 428L / 434S is the same Fc variant as 434S / 428L). For all positions discussed herein relating to antibodies or their derivatives and fragments (e.g., Fc domains), unless otherwise noted, amino acid numbering follows the EU index. The “EU index,” “EU index in Kabat,” or “EU numbering” scheme refers to the numbering of EU antibodies (Edelman et al., 1969, Proc Natl Acad Sci USA 63:78-85 (incorporated herein by reference)). Modifications may be additions, deletions, or substitutions.

[0105] Typically, a variant Fc domain has at least approximately 80, 85, 90, 95, 97, 98, or 99 percent identity with the corresponding parental human IgG Fc domain (using the identity algorithms discussed below (one embodiment utilizes the BLAST algorithm as known in the art using default parameters)). Alternatively, a variant Fc domain may have 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 or more amino acid modifications compared to the parental Fc domain. Alternatively, the variant Fc domain may have up to 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 11, 12, 13, 14, 15, 16, 17, 18, 19, or 20 amino acid modifications compared to the parent Fc domain. Furthermore, as discussed herein, the variant Fc domains described herein still retain the ability to form dimers with other Fc domains when measured using known techniques described herein, such as non-denaturing gel electrophoresis.

[0106] As used herein, "protein" means at least two covalently linked amino acids, and includes proteins, polypeptides, oligopeptides, and peptides. Furthermore, polypeptides constituting the antibodies of the present invention may include synthetic derivatization, glycosylation, PEGylation, cyclic substitution, cyclization, linking to other molecules, fusion to proteins or protein domains, and addition of peptide tags or labels to one or more side chains or terminals.

[0107] As used herein, "residue" means the position in a protein and its associated amino acid identity. For example, asparagine 297 (also referred to as Asn297 or N297) is the residue at position 297 in the human antibody IgG1.

[0108] As used herein, "IgG subclass modification" or "isotype modification" refers to an amino acid modification that converts a certain amino acid in one IgG isotype to a corresponding amino acid in a differently aligned IgG isotype. For example, since IgG1 contains tyrosine at EU position 296 and IgG2 contains phenylalanine, the F296Y substitution in IgG2 is considered an IgG subclass modification.

[0109] As used herein, “modifications not found in nature” means non-isomorphic amino acid modifications. For example, since none of the human IgGs contain serine at position 434, the substitution 434S in IgG1, IgG2, IgG3, or IgG4 (or their hybrids) is considered a modification not found in nature.

[0110] As used herein, "amino acid" and "amino acid identity" mean one of the 20 naturally occurring amino acids encoded by DNA and RNA.

[0111] As used herein, "effector function" means a biochemical event resulting from the interaction between an antibody Fc region and an Fc receptor or ligand. Effector functions include, but are not limited to, ADCC, ADCP, and CDC.

[0112] As used herein, "IgG Fc ligand" means a molecule, preferably a polypeptide, from any organism that binds to the Fc region of an IgG antibody to form an Fc / Fc ligand complex. Fc ligands include, but are not limited to, FcγRI, FcγRII, FcγRIII, FcRn, C1q, C3, mannan-binding lectins, mannose receptors, staphylococcal protein A, streptococcal protein G, and viral FcγR. Fc ligands also include Fc receptor homologs (FcRH), a family of Fc receptors homologous to FcγR (Davis et al., 2002, Immunological Reviews 190:123-136 (integrated as a whole by reference)). Fc ligands may include undiscovered molecules that bind Fc. Specific IgG Fc ligands are FcRn and Fc gamma receptors. As used herein, "Fc ligand" means a molecule, preferably a polypeptide, from any organism that binds to the Fc region of an antibody in order to form an Fc / Fc ligand complex.

[0113] When used herein, “Fc gamma receptor,” “FcγR,” or “Fc gamma R” means any member of the family of proteins that ligate to the Fc region of an IgG antibody and are encoded by the FcγR gene. In humans, this family includes, but is not limited to, FcγRI(CD64) including isoforms FcγRIa, FcγRIb, and FcγRIc; FcγRII(CD32) including isoform FcγRIIa (including allotypes H131 and R131), FcγRIIb (including FcγRIIb-1 and FcγRIIb-2), and FcγRIIc; and FcγRIII(CD16) including isoforms FcγRIIIa (including allotypes V158 and F158) and FcγRIIIb (including allotypes FcγRIIb-NA1 and FcγRIIb-NA2) (Jefferis et al., 2002, Immunol Lett 82:57-65 (incorporated by reference as a whole)), as well as any undiscovered human FcγR or FcγR isoform or allotype. FcγR can originate from any organism, including but not limited to humans, mice, rats, rabbits, and monkeys. Mouse FcγR includes, but is not limited to, FcγRI(CD64), FcγRII(CD32), FcγRIII(CD16), and FcγRIII-2(CD16-2), as well as any undiscovered mouse FcγR or FcγR isoform or allotype.

[0114] When used herein, “FcRn” or “fetal Fc receptor” means a protein that conjugates the Fc region of an IgG antibody and is at least partially encoded by the FcRn gene. FcRn may be from any organism, including but not limited to humans, mice, rats, rabbits, and monkeys. As is known in the art, a functional FcRn protein comprises two polypeptides, often referred to as a heavy chain and a light chain. The light chain is beta-2-microglobulin, and the heavy chain is encoded by the FcRn gene. Unless otherwise described herein, FcRn or FcRn protein refers to the complex of the FcRn heavy chain and beta-2-microglobulin. Various FcRn variants are used to increase binding to the FcRn receptor and, in some cases, to increase the serum half-life. “FcRn variant” is an amino acid modification that contributes to increased binding to the FcRn receptor, and preferred FcRn variants are shown below.

[0115] Where used herein, “parent polypeptide” means the initiating polypeptide that is later modified to produce a variant. The parent polypeptide may be a naturally occurring polypeptide, or a variant or engineered version of a naturally occurring polypeptide. Thus, where used herein, “parent immunoglobulin” means an unmodified immunoglobulin polypeptide that is modified to produce a variant, and “parent antibody” means an unmodified antibody that is modified to produce a variant antibody. It should be noted that “parent antibody” includes known commercially available recombinantly produced antibodies, as outlined below. In this context, “parent Fc domain” is relative to the described variant; thus, “variant human IgG1 Fc domain” is compared to the parent Fc domain of human IgG1, “variant human IgG4 Fc domain” is compared to the parent Fc domain of human IgG4, and so on.

[0116] As used herein, "position" means a location in the sequence of a protein. Positions may be numbered sequentially or in an established format, for example, according to the EU index for numbering antibody domains (e.g., CH1, CH2, CH3, or hinge domains).

[0117] As used herein, "target antigen" means a molecule that is specifically bound by an antigen-binding domain, which includes a variable region of a given antibody.

[0118] In the context of the monomers of the heterodimer antibodies of the present invention as used herein, “strandiness” means maintaining the ability to “pair” to form a heterodimer by incorporating the heterodimerizing variant into each monomer, similar to how two strands of DNA “pair.” For example, when manipulating monomer A with a certain pI variant (e.g., to increase the pI), a stereovariant that is a “charge pair” and can be similarly utilized does not interfere with the pI variant, and for example, a charge variant that increases the pI will be placed on the same “strand” or “monomer” to maintain the functionality of both. Similarly, with regard to “skew” variants provided in pairs of sets, as will be more thoroughly outlined below, those skilled in the art will consider the pI when determining which strand or monomer one set of the pair will be placed on, so that pI separation is also maximized using the pI of the skew.

[0119] As used herein, "target cell" means a cell that expresses a target antigen.

[0120] In relation to generating the bispecific antibody according to the present invention, "host cell" means a cell that contains exogenous nucleic acids encoding the components of the bispecific antibody and is capable of expressing the bispecific antibody under suitable conditions. Suitable host cells are discussed below.

[0121] In this specification, “wild-type” or “WT” means a naturally occurring amino acid or nucleotide sequence, including allele variations. WT proteins have an amino acid or nucleotide sequence that is not intentionally modified.

[0122] This specification provides numerous antibody domains (e.g., Fc domains) that have sequence identity with human antibody domains. Sequence identity between two similar sequences (e.g., antibody variable domains) can be determined using the following methods: Smith, TF & Waterman, MS (1981) “Comparison Of Biosequences,” Adv.Appl.Math.2:482 [Local homology algorithm]; Needleman, SB & Wunsch, CD (1970) “A General Method Applicable To The Search For Similarities In The Amino Acid Sequence Of Two Proteins,” J.Mol.Biol.48:443 [Homologous alignment algorithm]; Pearson, WR & Lipman, DJ (1988) “Improved Tools For Biological Sequence Comparison,” Proc.Natl.Acad.Sci.(USA)85:2444 [Similarity search method]; or Altschul, SF et al, (1990) “Basic Local Alignment Search Tool,” Sequence identity can be measured by algorithms such as the "BLAST" algorithm (see https: / / blast.ncbi.nlm.nih.gov / Blast.cgi), as described in J.Mol.Biol.215:403-10. When using any of the aforementioned algorithms, default parameters (regarding window length, gap penalty, etc.) are used. In one embodiment, sequence identity is performed using the BLAST algorithm with default parameters.

[0123] The antibodies of the present invention are typically isolated or recombinant. "Isolated", when used to describe the various polypeptides disclosed herein, means a polypeptide that has been identified, separated and / or recovered from the cells or cell culture in which it was expressed. Usually, an isolated polypeptide will be prepared by at least one purification step. "Isolated antibody" refers to an antibody that substantially does not contain other antibodies having different antigen specificities. "Recombinant" means that the antibodies are produced using recombinant nucleic acid techniques in exogenous host cells and that they can also be isolated.

[0124] "Specific binding" or "specifically binds to" or "specific for" a particular antigen or epitope means a binding that is clearly distinct from non-specific interactions. Specific binding can be measured, for example, by determining the binding of a molecule as compared to the binding of a control molecule, which is usually a molecule of similar structure that does not have binding activity. For example, specific binding can be determined by competition with a control molecule similar to the target.

[0125] Specific binding to a particular antigen or epitope is at least about 10 -4 M, at least about 10 -5 M, at least about 10 -6 M, at least about 10 -7 M, at least about 10 -8 M, at least about 10 -9 M, alternatively at least about 10 -10 M, at least about 10 -11 M, at least about 10 -12 M, or can be shown by an antibody having a KD for an antigen or epitope greater than that, where KD refers to the dissociation rate of a particular antibody-antigen interaction. Typically, an antibody that specifically binds an antigen will have a KD that is 20, 50, 100, 500, 1000, 5,000, 10,000 times higher or more than that for the control molecule with respect to the antigen or epitope.

[0126] Furthermore, specific binding to a particular antigen or epitope can be demonstrated by antibodies with a KA or Ka constant for the antigen or epitope that is at least 20, 50, 100, 500, 1,000, 5,000, or more than 10,000 times compared to the control, where KA or Ka refers to the binding constant of a particular antibody-antigen interaction. Binding affinity is typically measured using Biacore, SPR, or BLI assays.

[0127] IV. Anti-CD28 x anti-TROP2 antibody In one embodiment, a novel anti-CD28 × anti-TROP2 antibody is provided herein. In some embodiments, the anti-CD28 × anti-TROP2 antibody described herein can agonist-conjugate to the CD28 costimulatory molecule on T cells and to TROP2 on tumor cells. Such an antibody selectively enhances antitumor activity at TROP2-related tumor sites while minimizing peripheral toxicity. The target antibody provided herein is particularly useful in combination with other anticancer therapies, for example, a bispecific antibody for the treatment of TROP2-related cancers.

[0128] The anti-CD28 × anti-TROP2 antibody is polyvalent and comprises at least two antigen-binding domains (ABDs), where at least one antigen-binding domain is a CD28-binding domain and at least one antigen-binding domain is a TROP2-binding domain. The anti-CD28 × anti-TROP2 antibody of interest may contain any suitable CD28-binding domains and TROP2-binding domains, including, for example, the CD28-binding domains and TROP2-binding domains provided herein.

[0129] The antigen-binding domains provided herein generally include variable heavy domains (VH) having VH-CDR1, VH-CDR-2, and VH-CDR-3, and variable light domains (VL) having VL-CDR1, VL-CDR-2, and VL-CDR-3.

[0130] Furthermore, as discussed above, the numbering used to identify CDRs in sequence listings and figures is Kabat, but as shown in Table 2, different numbering can be used, which changes the amino acid sequence of the CDR.

[0131] Further variants can be created for all of the variable weight domains and variable light domains listed herein. As outlined herein, in some embodiments, a set of six CDRs may have 0, 1, 2, 3, 4, or 5 amino acid modifications (including amino acid substitutions used for specific applications), in addition to variations in the framework regions of the variable weight domains and variable light domains, provided that the framework (excluding the CDRs) retains at least approximately 80, 85, 90, 95, or 99% identity with a human germline sequence selected from those listed in Figure 1 of U.S. Patent No. 7,657,380 (the figure and legend are incorporated in their entirety by reference). Thus, for example, identical CDRs described herein can be combined with different framework sequences of human germline sequences, provided that the framework regions retain at least 80, 85, or 90% identity with a human germline sequence selected from those listed in Figure 1 of U.S. Patent No. 7,657,380. Alternatively, a CDR may have amino acid modifications (for example, one, two, three, four, or five amino acid modifications in a set of CDRs (i.e., a CDR may be modified as long as the total number of changes in a set of six CDRs is less than six amino acid modifications, and any combination of CDRs to be modified is arbitrary, for example, one change in vlCDR1, two changes in vhCDR2, and zero changes in vhCDR3)) as well as changes in the framework region, provided that the framework region maintains at least 80, 85, or 95-99% identity with respect to a human germline sequence selected from those enumerated in Figure 1 of U.S. Patent No. 7,657,380).

[0132] As will be understood by those skilled in the art, any set of six CDR or VH and VL domains may be in scFv or Fab form, which is then appended to the heavy and light chain constant domains, the heavy chain constant domain containing variants (including within the Fc domain in addition to within the CH1 domain).

[0133] In addition, in embodiments in which the target antibody contains scFv, the scFv may be oriented from the N-terminus to the C-terminus of a VH-scFv linker-VL or a VL-scFv linker-VH. In some forms, one or more ABDs are generally Fabs containing a VH domain on one protein chain (generally as a component of the heavy chain) and a VL domain on another protein chain (generally as a component of the light chain). Exemplary scFv linkers used in the target antibody are shown in Figure 6.

[0134] In some embodiments, the anti-CD28 × anti-TROP2 antibody is a bispecific antibody. In some embodiments, the anti-CD28 × anti-TROP2 antibody is a bivalent antibody. In some embodiments, the anti-CD28 × anti-TROP2 antibody is a trivalent antibody. In some embodiments, the anti-CD28 × anti-TROP2 antibody is a bispecific bivalent antibody. In some embodiments, the anti-CD28 × anti-TROP2 antibody contains one CD28-binding domain and one TROP2-binding domain. In exemplary embodiments, the anti-CD28 × anti-TROP2 antibody is a bispecific trivalent antibody. In some embodiments, the anti-CD28 × anti-TROP2 antibody contains one CD28-binding domain and two TROP2-binding domains.

[0135] The anti-CD28 × anti-TROP2 antibodies provided herein may be any useful form, including, for example, standard immunoglobulins, as well as the “1+1 Fab-scFv-Fc,” “2+1 mAb-scFv,” and “2+1 Fab2-scFv-Fc” forms described herein (Figure 15). Additional useful forms include, but are not limited to, the “mAb-Fv,” “central-Fv,” “1-arm scFv-mAb,” “scFv-mAb,” “double scFv,” and “trident” forms provided herein (e.g., Figure 15). See also US20180127501A1 (which is incorporated herein by reference, particularly the relevant parts relating to antibody forms) (e.g., see Figure 2). In some embodiments, the anti-CD28 × anti-TROP2 antibody is a heterodimerized bispecific antibody containing a variant Fc domain having one of the heterodimerized scuba riant, pI variant, and / or attenuation variants described herein. For example, please refer to Figure 8.

[0136] Unless otherwise specified herein, the order of names in the antigen list does not confer structure. That is, in the anti-TROP2 × anti-CD28 1+1 Fab-scFv-Fc antibody, scFv may bind to either TROP2 or CD28. However, in some cases, the order indicates the structure.

[0137] The anti-CD28 × anti-TROP2 antibodies provided herein further comprise different antibody domains. As described herein and known in the art, the antibodies described herein comprise different domains within the heavy and light chains, which may overlap. These domains include, but are not limited to, Fc domains, CH1 domains, CH2 domains, CH3 domains, hinge domains, heavy constant domains (CH1-hinge-Fc domain or CH1-hinge-CH2-CH3), variable heavy domains, variable light domains, light constant domains, Fab domains, and scFv domains.

[0138] As shown herein, there are numerous suitable linkers that can be used to covalently link described domains (e.g., scFv, Fab, Fc domains, VH domains, VL domains, etc.) containing conventional peptide bonds, generated by recombinant technology (for use as either domain linkers or scFv linkers). An exemplary linker for linking domains of a target antibody is shown in Figure 7. In some embodiments, the linker peptide may primarily contain the following amino acid residues: Gly, Ser, Ala, or Thr. The linker peptide should have a length appropriate for linking two molecules in such a way that they assume the correct conformation to each other so that they retain the desired activity. In one embodiment, the linker is about 1 to 50 amino acids long, preferably about 1 to 30 amino acids long. In one embodiment, a linker of 1 to 20 amino acids long may be used, and about 5 to about 10 amino acids are utilized in some embodiments. Useful linkers include, for example, glycine-serine polymers, glycine-alanine polymers, alanine-serine polymers, and other flexible linkers, including (GS)n, (GSGGS)n (SEQ ID NO: 750), (GGGGS)n (SEQ ID NO: 751), and (GGGS)n (SEQ ID NO: 752) (wherein n is an integer of at least 1 (and generally 3 to 4)). Alternatively, a variety of non-proteinaceous polymers, including but not limited to polyethylene glycol (PEG), polypropylene glycol, polyoxyalkylene, or copolymers of polyethylene glycol and polypropylene glycol, may be used as linkers.

[0139] Other linker sequences may include any sequence of any length from the CL / CH1 domain, but not all residues of the CL / CH1 domain, for example, the first 5-12 amino acid residues of the CL / CH1 domain. Linkers may originate from immunoglobulin light chains, e.g., Cκ or Cλ. Linkers may originate from immunoglobulin heavy chains of any isotype, including, for example, Cγ1, Cγ2, Cγ3, Cγ4, Cα1, Cα2, Cδ, Cε, and Cμ. Linker sequences may also originate from other proteins, e.g., Ig-like proteins (e.g., TCR, FcR, KIR), hinge region-derived sequences, and other native sequences from other proteins.

[0140] In some embodiments, the linker is a “domain linker” used to ligate any two domains outlined herein together. For example, in the 2+1 Fab2-scFv-Fc form, there may be a domain linker that ligates the C-terminus of the CH1 domain of Fab to the N-terminus of scFv, and another arbitrary domain linker that ligates the C-terminus of scFv to the CH2 domain (although in many embodiments, a hinge is used as this domain linker). Any suitable linker may be used, but many embodiments utilize a glycine-serine polymer as a domain linker, including, for example, (GS)n, (GSGGS)n (SEQ ID NO: 750), (GGGGS)n (SEQ ID NO: 751), and (GGGS)n (SEQ ID NO: 752) (wherein n is an integer of at least 1 (and generally 3-4-5)), in addition to any peptide sequence that enables recombination of two domains having sufficient length and flexibility to allow each domain to retain its biological function. In some cases, charged domain linkers, as used in some embodiments of the scFv linker, may be used, taking into account the "strand characteristics" outlined below. An exemplary useful domain linker is shown in Figure 7.

[0141] In some embodiments, the linker is an scFv linker used to covalently link the VH and VL domains, as discussed herein. Often, the scFv linker is a charged scFv linker, some of which are shown in Figure 6. Provided herein, so as therein, are charged scFv linkers for facilitating separation in pI between a first monomer and a second monomer. That is, by incorporating either a positive or negative charged scFv linker (or both; in the case of a skeleton using scFv on different monomers), monomers containing a charged linker can alter pI without causing further alteration to the Fc domain. These charged linkers can be substituted within any scFv containing a standard linker. In this case as well, as will be understood by those skilled in the art, the charged scFv linker is used on the correct “chain” or monomer according to the desired alteration of pI. For example, to produce a 1+1 Fab-scFv-Fc heterodimer antibody as discussed herein, the original pI of the Fv region for each of the desired antigen-binding domains is calculated, one is selected to produce the scFv, and a positive or negative linker is selected depending on the pI. Charged domain linkers can also be used to increase the pI separation of the monomers of the present invention, and thus those shown in Figure 4 can be used in any embodiment of this specification in which a linker is utilized.

[0142] Exemplary target anti-CD28×anti-TROP2 antibodies are shown, for example, in Figures 30 and 31. During cell culture generation of the anti-CD28×anti-TROP2 antibodies provided herein, a C-terminal lysine residue or a C-terminal lysine and glycine residue may be cleaved from the heavy chain monomer, thereby resulting in a C-terminus "cut-off" variant. See, for example, Jiang et al., Journal of Pharmaceutical Sciences 105:2066-2072 (2016). Thus, in some embodiments provided herein, the anti-CD28×anti-TROP2 antibody is a variant of one of the anti-CD28×anti-TROP2 antibodies, and includes a deletion of a C-terminal lysine (-K) terminus or a C-terminal lysine and glycine (-GK) residue in one or both Fc domains of the anti-CD28×anti-TROP2 antibodies described herein. In some embodiments, the deletion is G446del and / or K447del (EU numbering).

[0143] The form of the anti-CD28 × anti-TROP2 antibody is described in further detail below.

[0144] A. CD28 binding domain The anti-CD28 × anti-TROP2 antibodies provided herein include at least one CD28-binding domain. The anti-CD28 × anti-TROP2 antibodies provided herein may include any preferred CD28-binding domain. In exemplary embodiments, the CD28-binding domain is an agonist CD28 ABD that advantageously provides T-cell costimulatory activity.

[0145] As will be understood by those skilled in the art, a suitable CD28-binding domain may include a set of six CDRs, either as the underlined CDRs shown in the figure, or as CDRs identified using other alignments within the variable weight (VH) domain and variable light domain (VL) domain sequences of the CD28-binding domain shown in Figures 16, 19, and 22, when different numbering schemes are used as described herein and shown in Table 2. Additional VH and VL sequences of exemplary CD28-binding domains that can be used in the target antibody are shown in Figures 17, 18, and 34. A suitable CD28 ABD may also include these sequences and the entire VH and VL sequences shown in the figure, used as scFv or Fab.

[0146] In one embodiment, the CD28 antigen-binding domain includes, but is not limited to, those shown in Figures 16, 19, 21, and 22, any six CDRs of any of the CD28-binding domains described herein (i.e., vhCDR1-3 and vlCDR1-3). In some embodiments, the CD28 ABD that binds to human CD28 is the following CD28 One of the ABDs: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1.71, CD28.3[CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28]_H0L0, TGN1412_H1L1, 341VL 34[CD28]_H1L1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9.3[CD28]_H0L0, hu9.3[CD28]_H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, and TN228[CD28]_H4L2 (Figures 16, 19, and 22). In some embodiments, the CD28 ABD includes a VH / VL pair selected from VH and VL shown in Figures 16-19, 22, and 34.

[0147] In addition to the parent CDR sets disclosed in the Figures and Sequence Listings that form the ABD for CD28, variant CD28 ABDs are provided herein that have a CDR containing at least one modification of the CD28 ABD CDRs disclosed herein (e.g., Figures 16, 19, and 22, and the Sequence Listings). In one embodiment, the CD28 ABD of the target anti-CD28 × anti-TROP2 antibody includes a set of six CDRs having 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 amino acid modifications, compared to the six CDRs of the CD28 ABDs described herein, including the Figures and Sequence Listings. In an exemplary embodiment, the CD28 ABD of the target anti-CD28 × anti-TROP2 antibody includes the following CD28 When compared to one of the six CDRs in ABD, it includes a set of six CDRs with 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 amino acid modifications: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1.71, CD28.3[CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28]_H0L0, TGN1412 _H1L1, 341VL34[CD28]_H1L1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9.3[CD28]_H0L0, hu9.3[CD28]_H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, and TN228[CD28]_H4L2 (Figures 16, 19, and 22). In certain embodiments, the CD28 of the target anti-CD28 × anti-TROP2 antibody ABD can bind to the CD28 antigen when measured by at least one of the following assays: Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) (the latter being particularly used in many embodiments). In certain embodiments, CD28 ABD can bind to human CD28 antigen (see Figure 1).

[0148] In some embodiments, the CD28 ABD of the anti-CD28 × anti-TROP2 antibody in question comprises six CDRs that are at least 90, 95, 97, 98, or 99% identical to six CDRs of the CD28 ABD described herein, including in the figures and sequence listings. In exemplary embodiments, the CD28 ABD of the anti-CD28 × anti-TROP2 antibody in question comprises the following CD28 Includes six CDRs that are at least 90, 95, 97, 98, or 99% identical to one of the six CDRs in ABD: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1.71, CD28.3[CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28]_H 0L0, TGN1412_H1L1, 341VL34[CD28]_H1L1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9.3[CD28]_H0L0, hu9.3[CD28]_H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, and TN228[CD28]_H4L2 (Figures 16, 19, and 22). In certain embodiments, CD28 ABD can bind to CD28 when measured by at least one of Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) assays (the latter being particularly used in many embodiments). In certain embodiments, CD28 ABD can bind to human CD28 antigen (see Figure 1).

[0149] In another exemplary embodiment, the CD28 ABD of the target anti-CD28 × anti-TROP2 antibody comprises one of the variable weight (VH) domains and variable light (VL) domains of the CD28 ABDs described herein, including in the figures and sequence listings. In the exemplary embodiment, the CD28 ABD is the following CD28 One of the ABDs: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1.71, CD28.3[CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28]_H0L0, TGN1412_H1L1, 341VL 34[CD28]_H1L1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9.3[CD28]_H0L0, hu9.3[CD28]_H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, and TN228[CD28]_H4L2 (Figures 16, 19, and 22). In some embodiments, the CD28 ABD includes pairs of VH and VL selected from VH and VL shown in Figures 16-19, 22, and 34.

[0150] In some embodiments, the anti-CD28 × anti-TROP2 antibody comprises a CD28 ABD including a variable heavy domain and / or variable light domain, which are variants of the CD28 ABD VH and VL domains disclosed herein. In one embodiment, the variant VH domain and / or VL domain has 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the VH and / or VL domains of the CD28 ABD described herein, including figures and sequence listings. In exemplary embodiments, the variant VH domain and / or VL domain is the following CD28 One of the ABD has 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the VH and / or VL domain: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1.71, CD28.3[CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28] _H0L0, TGN1412_H1L1, 341VL34[CD28]_H1L1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9.3[CD28]_H0L0, hu9.3[CD28]_H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, and TN228[CD28]_H4L2 (Figures 16, 19, and 22). In some embodiments, the variation is located within the VH domain shown in Figures 16-19 and 22. In some embodiments, the changes are located within the VL domain shown in Figures 16-19 and 22. In some embodiments, the changes are located within the VH and VL domains shown in Figures 16-19 and 22. In some embodiments, one or more amino acid changes are located within the VH and / or VL framework regions (FR1, FR2, FR3, and / or FR4). In some embodiments, one or more amino acid changes are located within one or more CDRs.In certain embodiments, CD28 ABD can bind to CD28 when measured by at least one of Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) assays (the latter being particularly used in many embodiments). In certain embodiments, CD28 ABD can bind to human CD28 antigen (see Figure 1).

[0151] In one embodiment, the variant VH and / or VL domains are at least 90, 95, 97, 98, or 99% identical to the VH and / or VL of CD28 ABD described herein, including in the figures and sequence listings. In an exemplary embodiment, the variant VH and / or VL domains are the following CD28 These are at least 90, 95, 97, 98, or 99% identical to one of the VH and / or VL of ABD: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1.71, CD28.3[CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28]_H0L0 , TGN1412_H1L1, 341VL34[CD28]_H1L1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9.3[CD28]_H0L0, hu9.3[CD28]_H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, and TN228[CD28]_H4L2 (Figures 16, 19, and 22). In some embodiments, the CD28 ABD contains VH which is at least 90, 95, 97, 98, or 99% identical to the VH domain shown in Figures 16-19 and 22. In some embodiments, the CD28 ABD contains VL which is at least 90, 95, 97, 98, or 99% identical to the VL domain shown in Figures 16-19 and 22. In some embodiments, the CD28 ABD contains VH and VL which are at least 90, 95, 97, 98, or 99% identical to the VH and VL domains shown in Figures 16-19 and 22. In certain embodiments, the CD28 ABD can bind to CD28 when measured by at least one of Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) assays (the latter being particularly utilized in many embodiments).In certain embodiments, the CD28 ABD can bind to the human CD28 antigen (see Figure 1).

[0152] In some embodiments, the CD28-binding domain of the target anti-CD28 × anti-TROP2 antibody includes VH, which contains one of the sequences VHCDR1-3 or HFR1-4 shown in Figure 20A. In some embodiments, the CD28-binding domain includes VL, which contains one of the sequences VLCDR1-3 or LFR1-4 shown in Figure 20B.

[0153] In some embodiments, the anti-CD28 × anti-TROP2 antibody includes a CD28-binding domain comprising VH and VL selected from the following:

[0154] (i) VH containing vhCDR1, vhCDR2, and vhCDR3, wherein each VH has the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of any VH or variant thereof shown in Figures 16, 19, and 22; and (ii) VL containing vlCDR1, vlCDR2, and vlCDR3, wherein each VL has the amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of any VL or variant thereof shown in Figures 16, 19, and 22; or

[0155] (i) VH containing vhCDR1, vhCDR2, and vhCDR3, each having the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of the VH or its variant shown in Figure 17, and (ii) VL containing vlCDR1, vlCDR2, and vlCDR3, each having the amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of the VL or its variant shown in Figure 18.

[0156] In some embodiments, the anti-CD28 × anti-TROP2 antibody includes a CD28-binding domain comprising VH and VL selected from the following:

[0157] (i) a VH having the amino acid sequence of any VH or a variant thereof of the CD28 binding domain shown in Figures 16, 19, and 22, and (ii) a VL having the amino acid sequence of any VL or a variant thereof of the CD28 binding domain shown in Figures 16, 19, and 22; or

[0158] (i) VH having the amino acid sequence of VH or its variant shown in Figure 17 or 34, and (ii) VL having the amino acid sequence of VL or its variant shown in Figure 18.

[0159] 1. Additional CD28 binding domains In some embodiments, the anti-CD28 × anti-TROP2 antibodies provided herein include a CD28-binding domain containing a common light chain, the common light chain can also be used as the light chain of the TROP2-binding domain. In some embodiments, in such anti-CD28 × anti-TROP2 antibodies containing a CD28-binding domain, the CD28-binding domain and the TROP2-binding domain each contain a common light chain. Exemplary anti-CD28 × anti-TROP2 antibody forms utilizing a CD28-binding domain containing a common light chain include the 1+1 common light chain (CLC) and 2+1 common light chain forms described herein.

[0160] In one embodiment, the CD28 antigen-binding domain includes vhCDR1-3 of any of the CD28 variable heavy domains shown in Figure 34 and vlCDR1-3 of the 1F11-1A3.315 L1 variable light domain (see Figure 36A).

[0161] In one embodiment, the CD28 ABD of the target anti-CD28 × anti-TROP2 antibody includes a) vhCDR1, vhCDR2, and / or vhCDR3 having 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 amino acid modifications, respectively, when compared to one of the CD28 variable heavy domains shown in Figure 34, and / or vhCDR1, vhCDR2, and / or vhCDR3, respectively, when compared to vlCDR1, vlCDR2, and / or vlCDR3 of the 1F11-1A3.315 L1 variable light domain (see Figure 36A). In certain embodiments, the CD28 ABD of the target anti-CD28 × anti-TROP2 antibody is capable of binding to the CD28 antigen when measured by at least one of the following assays: Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) (the latter being particularly utilized in many embodiments). In certain embodiments, the CD28 ABD is capable of binding to the human CD28 antigen (see Figure 1).

[0162] In one embodiment, the CD28 ABD of the target anti-CD28 × anti-TROP2 antibody includes vhCDR1, vhCDR2, and / or vhCDR3 which are at least 90, 95, 97, 98, or 99% identical to one of the CD28 variable heavy domains vhCDR1, vhCDR2, and / or vhCDR3 shown in Figure 34, and / or vlCDR1, vlCDR2, and / or vlCDR3 which are at least 90, 95, 97, 98, or 99% identical to vlCDR1, vlCDR2, and / or vlCDR3 of the 1F11-1A3.315 L1 variable light domain (see Figure 36A). In certain embodiments, CD28 ABD can bind to CD28 when measured by at least one of Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) assays (the latter being particularly used in many embodiments). In certain embodiments, CD28 ABD can bind to human CD28 antigen (see Figure 1).

[0163] In another exemplary embodiment, the CD28 ABD of the target anti-CD28 × anti-TROP2 antibody includes one variable weight (VH) domain of the CD28 variable weight domains shown in Figure 34 and a 1F11-1A3.315 L1 variable light domain (see Figure 36A).

[0164] In some embodiments, the anti-CD28 × anti-TROP2 antibody contains a CD28 ABD comprising a variable heavy domain which is a variant of one of the CD28 variable heavy domains shown in Figure 34 and / or a variable light domain which is a variant of the 1F11-1A3.315 L1 variable light domain (see Figure 36A). In some embodiments, the variant VH domain has 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes compared to one of the CD28 variable heavy domains shown in Figure 34, and / or the variable light domain has 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the 1F11-1A3.315 L1 variable light domain (see Figure 36A). In some embodiments, one or more amino acid changes are located within the VH and / or VL framework regions (FR1, FR2, FR3, and / or FR4). In some embodiments, one or more amino acid changes are located within one or more CDRs. In certain embodiments, CD28 ABD can bind to CD28 when measured by at least one of Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) assays (the latter being particularly used in many embodiments). In certain embodiments, CD28 ABD can bind to human CD28 antigen (see Figure 1).

[0165] In one embodiment, the variant VH domain is at least 90, 95, 97, 98, or 99% identical to one of the CD28 variable heavy domains shown in Figure 34, and / or the variable light domain is at least 90, 95, 97, 98, or 99% identical to the 1F11-1A3.315 L1 variable light domain (see Figure 36A). In certain embodiments, the CD28 ABD can bind to CD28 when measured by at least one of Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) assays (the latter being particularly used in many embodiments). In certain embodiments, the CD28 ABD can bind to the human CD28 antigen (see Figure 1).

[0166] In some embodiments, the CD28-binding domain of the target anti-CD28 × anti-TROP2 antibody contains a VH including one of the VHCDR1-3 or HFR1-4 sequences shown in Figure 34C. In some embodiments, the CD28-binding domain contains one of the VLCDR1-3 or LFR1-4 sequences of the 1F11-1A3.315 L1 variable light domain (see Figure 36A).

[0167] B-TROP2 binding domain The anti-CD28 × anti-TROP2 antibodies provided herein include at least one TROP2-binding domain. A target antibody containing such a TROP2 antigen-binding domain (e.g., an anti-TROP2 × anti-CD3 bispecific antibody) favorably targets cells expressing higher levels of TROP2 than cells expressing TROP2 levels (e.g., normal cells).

[0168] As those skilled in the art will recognize, a preferred TROP2 binding domain includes a set of six CDRs, either as underlined CDRs as shown in the sequence listing and Figures 23-28, or as CDRs identified using other alignments within the variable weight (VH) domain and variable light domain (VL) sequences shown in Figures 23-28 and the sequence listing, when different numbering schemes are used as described herein and shown in Table 2 (see Table 2). A preferred TROP2 ABD may also include these sequences and the entire VH and VL sequences shown in the figures, used as scFv or Fab domains.

[0169] In exemplary embodiments, the TROP2 ABD of the anti-CD28 × anti-TROP2 antibody comprises one variable weight (VH) domain and a variable light (VL) domain of any of the TROP2 ABDs described herein, including figures and sequence listings. In exemplary embodiments, the TROP2 ABD is the TROP2 ABD shown in Figure 23. In exemplary embodiments, the TROP2 ABD comprises a pair of VH and VL domains selected from those shown in Figures 23–28.

[0170] In addition to the parent TROP2 variable heavy domain and variable light domain disclosed herein, TROP2 ABDs are provided herein that include variable heavy domains and / or variable light domains that are variants of the TROP2 ABD VH and VL domains disclosed herein. In one embodiment, the variant VH domain and / or VL domain has 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the VH and / or VL domains of TROP2 ABD described herein, including figures and sequence listings. In an exemplary embodiment, the variant VH domain and / or VL domain has 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the VH and / or VL domains of TROP2 ABD shown in Figure 23. In an exemplary embodiment, the variant VH domain and / or VL domain has 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the VH and / or VL domains shown in Figures 23-28. In some embodiments, the alteration is located within the VH domain shown in Figures 23-28. In some embodiments, the alteration is located within the VL domain shown in Figures 23-28. In some embodiments, the alteration is located within both the VH and VL domains shown in Figures 23-28. In some embodiments, one or more amino acid alterations are located within the VH and / or VL framework regions (FR1, FR2, FR3, and / or FR4). In some embodiments, one or more amino acid alterations are located within one or more of the vhCDR1-3 and / or vlCDR1-3. In certain embodiments, the TROP2 ABD of the anti-CD28 × anti-TROP2 antibody is capable of binding to TROP2 when measured by at least one of the following assays: Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) assay (the latter being particularly used in many embodiments). In certain embodiments, the TROP2 ABD is capable of binding to the human TROP2 antigen (Figure 2).

[0171] In one embodiment, the variant VH and / or VL domains are at least 90, 95, 97, 98, or 99% identical to the VH and / or VL of TROP2 ABD described herein, including in the figures and sequence listings. In an exemplary embodiment, the variant VH and / or VL domains are at least 90, 95, 97, 98, or 99% identical to the VH and / or VL of TROP2 ABD shown in Figure 23. In an exemplary embodiment, the variant VH and / or VL domains are at least 90, 95, 97, 98, or 99% identical to the VH and / or VL shown in Figures 23-28. In some embodiments, TROP2 ABD includes a VH that is at least 90, 95, 97, 98, or 99% identical to the VH domain shown in Figures 23-28. In some embodiments, the TROP2 ABD includes a VL that is at least 90, 95, 97, 98, or 99% identical to the VL domain shown in Figures 23-28. In some embodiments, the TROP2 ABD includes VH and VL that are at least 90, 95, 97, 98, or 99% identical to the VH and VL domains shown in Figures 23-28. In certain embodiments, the TROP2 ABD of an anti-CD28 × anti-TROP2 antibody is capable of binding to TROP2 when measured by at least one of Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) assays (the latter being particularly utilized in many embodiments). In certain embodiments, the TROP2 ABD is capable of binding to the human TROP2 antigen (Figure 2).

[0172] In some embodiments, the anti-CD28 × anti-TROP2 antibody is a bivalent antibody (e.g., a 1+1 Fab-scFv-Fc antibody) containing one TROP2-binding domain. In other embodiments, the anti-CD28 × anti-TROP2 antibody is a trivalent antibody (e.g., a 2+1 mAb-scFv and 2+1 Fab2-scFv-Fc antibody) containing two TROP2-binding domains.

[0173] 1. Additional TROP2 binding domains This specification provides TROP2 binding domain compositions comprising a common variable light domain, the common variable light domain may also be advantageously included in the CD28 and PDL1 binding domains described herein. In some embodiments, the TROP2 antigen-binding domain comprises one of the TROP2 variable heavy domains shown in Figure 36 and the 1F11-1A3.315 L1 variable light domain (see Figure 36A, also referred to herein as the "IGKV1-39" variable light domain). In some embodiments, the TROP2 antigen-binding domain comprises one of the TROP2 antigen-binding domains shown in Figures 52-56, a variable heavy domain and a variable light domain or a variant thereof.

[0174] In some embodiments, the anti-CD28 × anti-TROP2 antibodies provided herein include a TROP2-binding domain containing a common light chain, the common light chain of which can also be used as the light chain of the CD28-binding domain. In some embodiments, in such anti-CD28 × anti-TROP2 antibodies containing a TROP2-binding domain, the CD28-binding domain and the TROP2-binding domain each contain a common light chain. Exemplary anti-CD28 × anti-TROP2 antibody forms utilizing a TROP2-binding domain containing a common light chain include the 1+1 common light chain (CLC) and 2+1 common light chain forms described herein.

[0175] In one embodiment, the TROP2 antigen-binding domain includes vhCDR1-3 of any of the TROP2 variable heavy domains shown in Figure 36 and vlCDR1-3 of the 1F11-1A3.315 L1 variable light domain (see Figure 36A). In some embodiments, the TROP2 antigen-binding domain includes vhCDR1-3 and vlCDR1-3 of any one of the TROP2 antigen-binding domains shown in Figures 52-56.

[0176] In one embodiment, the TROP2 ABD of the target anti-CD28 × anti-TROP2 antibody includes a) vhCDR1, vhCDR2, and / or vhCDR3 having 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 amino acid modifications, respectively, when compared to one of the TROP2 variable heavy domains shown in Figures 36 and 52-56, and / or b) vlCDR1, vlCDR2, and / or vlCDR3 having 1, 2, 3, 4, 5, 6, 7, 8, 9, and 10 amino acid modifications, respectively, when compared to the vlCDR1, vlCDR2, and / or vlCDR3 of the 1F11-1A3.315 L1 variable light domain (see Figures 36A and 52-56). In certain embodiments, the TROP2 ABD of the target anti-CD28 × anti-TROP2 antibody is capable of binding to the TROP2 antigen when measured by at least one of Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) assays (the latter being particularly utilized in many embodiments). In certain embodiments, the TROP2 ABD is capable of binding to the human TROP2 antigen (see Figure 2).

[0177] In one embodiment, the TROP2 ABD of the target anti-CD28 × anti-TROP2 antibody includes vhCDR1, vhCDR2, and / or vhCDR3 which are at least 90, 95, 97, 98, or 99% identical to one of the TROP2 variable heavy domains vhCDR1, vhCDR2, and / or vhCDR3 shown in Figures 36 and 52-56, respectively, and vlCDR1, vlCDR2, and / or vlCDR3 which are at least 90, 95, 97, 98, or 99% identical to vlCDR1, vlCDR2, and / or vlCDR3 of the 1F11-1A3.315 L1 variable light domain (see Figures 36A and 52-56). In certain embodiments, TROP2 ABD can bind to the TROP2 antigen when measured by at least one of the following assays: Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interference assay, e.g., Octet assay) (the latter being particularly utilized in many embodiments). In certain embodiments, TROP2 ABD can bind to human TROP2 antigen (see Figure 2).

[0178] In another exemplary embodiment, the TROP2 ABD of the anti-CD28×anti-TROP2 antibody in question includes one variable weight (VH) domain from the TROP2 variable weight domains shown in Figures 36 and 52-56, and a 1F11-1A3.315 L1 variable light domain (see Figures 36A and 52-56).

[0179] In some embodiments, the anti-CD28 × anti-TROP2 antibody comprises a TROP2 ABD including a variable heavy domain which is a variant of one of the TROP2 variable heavy domains shown in Figures 36 and 52-56, and / or a variable light domain which is a variant of the 1F11-1A3.315 L1 variable light domain (see Figures 36A and 52-56). In some embodiments, the variant VH domain has 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes compared to one of the TROP2 variable heavy domains shown in Figures 36 and 52-56, and / or the variable light domain has 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 amino acid changes from the 1F11-1A3.315 L1 variable light domain (see Figures 36A and 52-56). In some embodiments, one or more amino acid changes are located within the VH and / or VL framework regions (FR1, FR2, FR3, and / or FR4). In some embodiments, one or more amino acid changes are present in one or more CDRs. In certain embodiments, the TROP2 ABD can bind to TROP2 when measured by at least one of the following assays: Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interference, e.g., Octet assay) (the latter being particularly used in many embodiments). In certain embodiments, the TROP2 ABD can bind to the human TROP2 antigen (see Figure 2).

[0180] In one embodiment, the variant VH domain is at least 90, 95, 97, 98, or 99% identical to one of the TROP2 variable heavy domains shown in Figures 36 and 52-56, and / or the variable light domain is at least 90, 95, 97, 98, or 99% identical to the 1F11-1A3.315 L1 variable light domain (see Figures 36A and 52-56). In certain embodiments, the TROP2 ABD can bind to TROP2 when measured by at least one of Biacore, surface plasmon resonance (SPR), flow cytometry, and / or BLI (biolayer interferometry, e.g., Octet assay) assays (the latter being particularly used in many embodiments). In certain embodiments, the TROP2 ABD can bind to the human TROP2 antigen (see Figure 2).

[0181] In some embodiments, the TROP2-binding domain of the target anti-CD28 × anti-TROP2 antibody contains a VH including one of the VHCDR1-3 or HFR1-4 sequences shown in Figure 36C. In some embodiments, the TROP2-binding domain contains one of the VLCDR1-3 or LFR1-4 sequences of the 1F11-1A3.315 L1 variable light domain (see Figure 36A).

[0182] C. Chimeric antibodies and humanized antibodies In certain embodiments, the target antibodies provided herein include heavy chain variable regions from specific germline heavy chain immunoglobulin genes and / or light chain variable regions from specific germline light chain immunoglobulin genes. For example, such antibodies include, or may consist of, human antibodies containing heavy chain or light chain variable regions that are "products of" or "derived from" a specific germline sequence. Human antibodies that are "products of" or "derived from" a human germline immunoglobulin sequence can be identified, for example, by comparing the amino acid sequence of the human antibody with the amino acid sequence of the human germline immunoglobulin and selecting the human germline immunoglobulin sequence that is closest in sequence to (i.e., maximum % identity) the sequence of the human antibody (using the methods outlined herein). Human antibodies that are "products of" or "derived from" a specific human germline immunoglobulin sequence may contain amino acid differences when compared to the germline sequence, for example, by the intentional introduction of naturally occurring somatic mutations or site-directed mutations. However, humanized antibodies are typically at least 90% identical to the amino acid sequence encoded by human germline immunoglobulin genes and contain amino acid residues that identify the antibody as derived from a human sequence when compared to germline immunoglobulin amino acid sequences of other species (e.g., mouse germline sequences). In given cases, a humanized antibody may be at least 95, 96, 97, 98, or 99% identical in amino acid sequence to the amino acid sequence encoded by a germline immunoglobulin gene, or even at least 96%, 97%, 98, or 99%. Typically, a humanized antibody derived from a particular human germline sequence will show a difference of 10 to 20 amino acids or less from the amino acid sequence encoded by the human germline immunoglobulin gene (before the introduction of any skew, pI, and attenuation variants as used herein (i.e., the number of variants is usually small before the introduction of the variants of the present invention)).In certain cases, the humanized antibody may exhibit a difference of 5 or fewer amino acids, or further 4, 3, 2, or 1 or fewer, from the amino acid sequence encoded by the germline immunoglobulin gene (in this case as well, before the introduction of any skew, pI, and attenuation variants as specified herein (i.e., the number of variants is usually small before the introduction of the variants of the present invention)).

[0183] In one embodiment, the parent antibody is affinity-matured as known in the art. Structure-based methods may be used for humanization and affinity maturation, for example, as described in USSN 11 / 004,590. Humanization and / or affinity maturation of the antibody variable region may be performed using selection-based methods, including, but not limited to, those described in Wu et al., 1999, J.Mol.Biol.294:151-162; Baca et al., 1997, J.Biol.Chem.272(16):10678-10684; Rosok et al., 1996, J.Biol.Chem.271(37):22611-22618; Rader et al., 1998, Proc.Natl.Acad.Sci.USA 95:8910-8915; Krauss et al., 2003, Protein Engineering 16(10):753-759 (all incorporated by reference). Other humanization methods, including but not limited to those described in USSN09 / 810,510; Tan et al.,2002,J.Immunol.169:1119-1125;De Pascalis et al.,2002,J.Immunol.169:3076-3084 (all incorporated by reference), may involve grafting only a portion of the CDR.

[0184] D. Heterodimer antibodies In exemplary embodiments, the anti-CD28×anti-TROP2 antibody provided herein is a heterodimer bispecific antibody comprising two variant Fc domain sequences. Such variant Fc domains include amino acid modifications to facilitate the self-construction and / or purification of the heterodimer antibody.

[0185] An ongoing challenge in antibody technology is the desire for "bispecific" antibodies that bind simultaneously to two different antigens, typically by bringing the different antigens into proximity and resulting in novel functionalities and therapies. These antibodies are usually produced by incorporating genes for each heavy and light chain into the host cell. This typically results in the formation of the desired heterodimer (AB), as well as two homodimers (AA and BB (without the light chain heterodimer problem)). However, a major obstacle in the formation of bispecific antibodies is the bias towards the formation of the desired heterodimer antibody relative to the formation of the homodimer, and / or the difficulty in purifying the heterodimer antibody from the homodimer.

[0186] Numerous mechanisms exist that can be used to generate target heterodimer antibodies. Furthermore, as will be understood by those skilled in the art, these different mechanisms can be combined to ensure high heterodimerization. Amino acid modifications that facilitate the generation and purification of heterodimers are usually collectively referred to as “heterodimerizing variants.” As will be discussed below, heterodimerizing variants include “skew” variants (e.g., the “knob and hole” and “charge pair” variants described below) as well as “pI variants” that enable the purification of heterodimers from homodimers. Useful mechanisms for heterodimerization, as described in U.S. Patent No. US9,605,084 (which in whole and in particular is incorporated herein by reference with respect to the discussion of heterodimerization variants below), include “knobs and holes” ("KIH") as described in U.S. Patent No. US9,605,084, “electrostatic steering” or “charge pairs” as described in U.S. Patent No. US9,605,084, pI variants as described in U.S. Patent No. US9,605,084, and additional general Fc variants outlined in U.S. Patent No. US9,605,084 and below.

[0187] Heterodimizing variants useful for the formation and purification of target heterodimer antibodies (e.g., bispecific antibodies) are discussed in more detail below.

[0188] 1. Scubarian In some embodiments, heterodimer antibodies include scubarians, which are one or more amino acid modifications in the first Fc domain (A) and / or the second Fc domain (B) that prioritize the formation of an Fc heterodimer (an Fc dimer (AB) containing the first and second Fc domains) rather than an Fc homodimer (an Fc dimer (AA or BB) containing two first Fc domains or two second Fc domains). Preferred scubarians are shown in Figure 29, and in Figures 3 and 9, of U.S. Publication No. 2016 / 0355608 (which is incorporated herein by reference in whole and in particular with respect to its disclosure of scubarians).

[0189] One particular type of scFv variant is sometimes referred to in the art as a "knob and hole" which refers to amino acid manipulations that create steric effects so as to prefer heterodimer formation and not homodimer formation, as described in USSN 61 / 596,846, Ridgway et al., Protein Engineering 9(7):617(1996); Atwell et al., J. Mol. Biol. 1997 270:26; U.S. Patent No. 8,216,805 (all of which are hereby incorporated by reference in their entirety and particularly for the disclosure of the "knob and hole" mutations). These are sometimes referred to herein as "steric variants". The figure identifies a number of "monomer A - monomer B" pairs that depend on the "knob and hole". Also, as described in Merchant et al., Nature Biotech. 16:677(1998), these "knob and hole" mutations can be combined with disulfide bonds to further prefer the formation of Fc heterodimers.

[0190] Another method utilized in the generation of heterodimers is sometimes referred to as "electrostatic steering", as described in Gunasekaran et al., J. Biol. Chem. 285(25):19637(2010) (which is hereby incorporated by reference in its entirety). This is sometimes referred to herein as a "charge pair". In this embodiment, electrostatics is used to bias the formation towards heterodimerization. As will be appreciated by those skilled in the art, these can also affect the pI and thus purification, and thus in some cases can also be considered pI variants. However, since these were generated to enforce heterodimerization and were not used as purification tools, they are classified as "scuba variants". These include, but are not limited to, D221E / P228E / L368E paired with D221R / P228R / K409R (e.g., these are "corresponding monomer sets") and C220E / P228E / 368E paired with C220R / E224R / P228R / K409R.

[0191] In some embodiments, scubarians favorably and simultaneously prioritize heterodimerization based on both a “knob and hole” mechanism and an “electrostatic steering” mechanism. In some embodiments, a heterodimer antibody contains one or more sets of such heterodimerized scubarians. These variants are provided in “pairs” of “sets,” that is, one set of the pair is incorporated into a first monomer, and the other set of the pair is incorporated into a second monomer. It should be noted that these sets do not necessarily behave as “knob in hole” variants having a one-to-one correspondence between residues on one monomer and residues on the other monomer. In other words, pairs of these sets can rather form an interface between the two monomers, thereby promoting heterodimerization and suppressing homodimerization, making it possible to raise the proportion of heterodimers spontaneously formed under biological conditions to over 90% (25% homodimer A / A: 50% heterodimer A / B: 25% homodimer B / B) rather than the expected 50%. Exemplary heterodimerization "skew" variants are shown in Figures 3 and 9. Such "skewed" variants include, but are not limited to, S364K / E357Q:L368D / K370S;L368D / K370S:S364K;L368E / K370S:S364K;T411T / E360E / Q362E:D401K;L368D / K370S:S364K / E357L;K370S:S364K / E357Q (EU numbering).

[0192] In exemplary embodiments, the heterodimer antibody includes the S364K / E357Q:L368D / K370S;L368D / K370S:S364K;L368E / K370S:S364K;T411T / E360E / Q362E:D401K;L368D / K370S:S364K / E357L;K370S:S364K / E357Q; or T366S / L368A / Y407V:T366W (optionally containing a cross-linked disulfide, T366S / L368A / Y407V / Y349C:T366W / S354C) "skew" variant amino acid substitution set (EU numbering). In exemplary embodiments, the heterodimer antibody contains the "S364K / E357Q:L368D / K370S" amino acid substitution set. From a nomenclature standpoint, the pair "S364K / E357Q:L368D / K370S" means that one monomer contains an Fc domain with amino acid substitutions S364K and E357Q, and the other monomer contains an Fc domain with amino acid substitutions L368D and K370S; as described above, the "strandiness" of these pairs depends on the initiation pI.

[0193] In some embodiments, the scuba riants provided herein may, optionally and independently, be incorporated into one or both of the first and second Fc domains of a heterodimer antibody along with any other modifications, including but not limited to other scuba riants (see, for example, Figure 37 of U.S. Publication No. 2012 / 0149876 (which is incorporated herein by reference in particular for its disclosure of scuba riants)), pI variants, isomorphic variants, FcRn variants, attenuation variants, and the like. Furthermore, individual modifications may also, independently and optionally, be included in or excluded from the heterodimer antibody in question.

[0194] In some embodiments, the scuba rians outlined herein may be optionally and independently incorporated into one or both heavy chain monomers together with any pI variant (or other variants, e.g., Fc variant, FcRn variant, etc.), and may be independently and optionally included in or excluded from the heterodimer antibody of interest.

[0195] 2. Purified variant In some embodiments, the heterodimer antibody includes a purified variant that advantageously enables the separation of the heterodimer protein (e.g., anti-CD28 × anti-TROP2 bispecific antibody) from the homodimer protein.

[0196] Several fundamental mechanisms exist that can facilitate the purification of heterodimer antibodies. For example, modifications to one or both of antibody heavy chain monomers A and B, such that each monomer has a different pI, enable isoelectric purification of heterodimer AB antibodies from monomer AA and BB proteins. Alternatively, several skeletal configurations, such as the "1+1 Fab-scFv-Fc" configuration and the "2+1 Fab2-scFv-Fc" configuration, enable size-based separation. As mentioned above, it is also possible to "distort" heterodimer formation relative to homodimers using scuba riants. Therefore, combinations of heterodimerizing scuba riants and purification variants are particularly useful for the heterodimer antibodies provided herein.

[0197] Furthermore, as will be outlined more thoroughly below, depending on the form of the heterodimeric antibody, purified variants contained in the constant region and / or Fc domain and / or domain linker of the monomer may be used. In some embodiments, the heterodimeric antibody includes additional modifications for alternative functionality, such as Fc, FcRn, and KO variants, which may also produce pI changes.

[0198] In some embodiments, the heterodimer antibodies provided herein include at least one monomer having one or more modifications (i.e., “pI variants”) that alter the pI of the monomer. Generally, as understood by those skilled in the art, there are two general categories of pI variants: those that increase the pI of the protein (basic changes) and those that decrease the pI of the protein (acidic changes). As described herein, all combinations of these variants are possible: one monomer may be wild-type, or a variant that does not exhibit a significantly different pI from the wild-type, while the other may be either more basic or more acidic. Alternatively, each monomer may be modified to be one more basic and the other more acidic.

[0199] Depending on the form of the heterodimer antibody, the pI variant may be contained within the constant and / or Fc domain of the monomer, or a charged linker (either a domain linker or an scFv linker) may be used. That is, antibody forms utilizing scFv(s), e.g., the "1+1 Fab-scFv-Fc" form, may include a charged scFv linker (either positive or negative) to provide a further increase in pI for purification purposes. As will be understood by those skilled in the art, some 1+1 Fab-scFv-Fc and 2+1 Fab2-scFv-Fc forms are useful simply with a charged scFv linker and without additional pI adjustment, but the present invention also provides pI variants located in one or both monomers and / or in the charged domain linker. Furthermore, additional amino acid manipulation for alternative functionality may also confer pI changes, e.g., Fc, FcRn, and KO variants.

[0200] In target heterodimer antibodies that utilize pI as a separation mechanism to enable the purification of heterodimer proteins, amino acid variants are introduced into one or both monomeric polypeptides. That is, the pI of one monomer (referred herein as "monomer A" for simplicity) may be manipulated separately from monomer B, or both monomer A and B changes may be altered by increasing pI for monomer A and decreasing pI for monomer B. As will be more thoroughly outlined below, pI changes in either or both monomers may be made by removing or adding charged residues (e.g., a neutral amino acid is replaced by a positively or negatively charged amino acid residue (e.g., glycine to glutamic acid)), changing a charged residue from positive or negative to the opposite charge (e.g., aspartic acid to lysine), or changing a charged residue to a neutral residue (e.g., loss of charge; lysine to serine). Numerous of these variants are shown in Figure 4.

[0201] Therefore, in some embodiments, the target heterodimer antibody includes amino acid modifications in the constant region that alter the isoelectric point (pI) of at least one (if not both) monomer of the dimer protein in order to form a “pI antibody” by incorporating an amino acid substitution (“pI variant” or “pI substitution”) in one or both monomers. As shown herein, if the pIs of the two monomers differ by only 0.1 pH units (0.2, 0.3, 0.4, and 0.5 or more are all utilized in the present invention), separation of heterodimers from two homodimers can be achieved.

[0202] As will be understood by those skilled in the art, the number of pI variants present in each or both monomers(s) to obtain good separation will depend in part on the starting pIs of the components, e.g., in the form 1+1 Fab-scFv-Fc, 2+1 Fab2-scFv-Fc, 1+1 CLC, and 2+1 CLC, on the scFv(1+1 Fab-scFv-Fc, 2+1 Fab2-scFv-Fc) and Fab(s)(s) of the target(s). That is, the Fv sequences of the two target antigens are calculated and then the decision is made to determine which monomers to operate on or in which "direction" (e.g., more positive or more negative). As is known in the art, different Fvs will have different starting pIs to utilize in the present invention. Typically, as outlined herein, the pIs are operated to result in a total pI difference of at least about 0.1 log (preferably 0.2 to 0.5 as outlined herein) between each monomer.

[0203] When pI variants are used to achieve heterodimerization, a more modular approach is provided for designing and purifying antibody-containing bispecific proteins by utilizing the constant region(s) of the heavy chain(s). Thus, in some embodiments, heterodimerizing variants (including skew and pI heterodimerizing variants) are not included in the variable region, so that each individual antibody must be manipulated. Also, in some embodiments, the potential for immunogenicity due to pI variants is significantly reduced by introducing pI variants from different IgG isotypes so that pI can be altered without introducing significant immunogenicity. Thus, an additional problem to be solved is the elucidation of low pI constant domains with a high human sequence content, e.g., minimizing or avoiding non-human residues at any particular position. As an alternative to or in addition to isomorphic substitutions, the potential for immunogenicity due to pI variants is significantly reduced by utilizing isosteric substitutions (e.g., Asn to Asp; and Gln to Glu).

[0204] As discussed below, the secondary benefits that may arise from this pI manipulation are also the extension of serum half-life and increased FcRn binding. Specifically, as described in U.S. Publication No. US2012 / 0028304 (which is incorporated in its entirety by reference), reducing the pI of antibody constant domains (including those found in antibodies and Fc fusions) can result in longer serum retention in vivo. These pI variants for increased serum half-life also facilitate pI alteration for purification.

[0205] Furthermore, it should be noted that pI variants provide additional benefits for the analysis and quality control processes of bispecific antibodies, as their ability to exclude, minimize, and distinguish homodimers is important. Similarly, the ability to reliably test the reproducibility of heterodimeric antibody production is important.

[0206] Typically, the most commonly used embodiments rely on a set of variants, combining a scuba riant that promotes heterodimerization rather than homodimerization with a pI variant that increases the pI difference between the two monomers, making it easier to purify the heterodimer from the homodimer.

[0207] Exemplary combinations of pI variants are shown in Figures 3 and 4, and Figure 30, of U.S. Publication No. 2016 / 0355608 (all of which are incorporated herein by reference in whole and in particular for the disclosure of pI variants). Preferred combinations of pI variants are shown in Figures 3 and 4. As outlined herein and shown in the figures, these variations are shown relative to IgG1, but all isotypes and isotype hybrids can be modified in the same way. R133E and R133Q can also be used when the heavy chain constant domain is derived from IgG2-4.

[0208] In one embodiment, a preferred combination of pI variants has one monomer (negative Fab side) containing the 208D / 295E / 384D / 418E / 421D variant (N208D / Q295E / N384D / Q418E / N421D compared to human IgG1) and a second monomer (positive scFv side) containing a positively charged scFv linker containing (GKPGS)4 (SEQ ID NO: 24). However, as will be understood by those skilled in the art, the first monomer contains a CH1 domain including position 208. Therefore, in constructs that do not contain the CH1 domain (for example, antibodies that do not utilize the CH1 domain in one of their domains), the preferred negative pI variant Fc set includes the 295E / 384D / 418E / 421D variants (Q295E / N384D / Q418E / N421D compared to human IgG1).

[0209] Therefore, in some embodiments, one monomer has the set of substitutions shown in Figure 8, and the other monomer has a charged linker (either a charged scFv linker (since the monomer contains scFv) or a charged domain linker (which can be selected from those shown in Figure 6, as the form indicates)).

[0210] In some embodiments, modifications are made at the hinge of the Fc domain, including positions 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229, and 230, based on EU numbering. Thus, pI mutations and especially substitutions may occur at one or more of positions 216-230, with 1, 2, 3, 4, or 5 mutations being utilized. In this case as well, all possible combinations are intended alone or in conjunction with other pI variants in other domains.

[0211] Specific substitutions utilized when lowering the pI of the hinge domain include, but are not limited to, a deletion at position 221, a non-native valine or threonine at position 222, a deletion at position 223, a non-native glutamic acid at position 224, a deletion at position 225, a deletion at position 235, and a deletion or non-native alanine at position 236. In some cases, only the pI substitution is made in the hinge domain, and in other cases, these substitution(s) may be added in any combination to other pI variants in other domains.

[0212] In some embodiments, the mutations can be made in the CH2 region, including positions 233, 234, 235, 236, 274, 296, 300, 309, 320, 322, 326, 327, 334, and 339, based on the EU numbering. It should be noted that the changes at 233 - 236 can be made (along with 327A) to increase effector function in the IgG2 backbone. Again, all possible combinations of these 14 positions can be made; for example, the antibodies provided herein can include variant Fc domains having 1, 2, 3, 4, 5, 6, 7, 8, 9, or 10 CH2 pI substitutions.

[0213] Specific substitutions utilized when lowering the pI of the CH2 domain include, but are not limited to, a non-native glutamine or glutamic acid at position 274, a non-native phenylalanine at position 296, a non-native phenylalanine at position 300, a non-native valine at position 309, a non-native glutamic acid at position 320, a non-native glutamic acid at position 322, a non-native glutamic acid at position 326, a non-native glycine at position 327, a non-native glutamic acid at position 334, a non-native threonine at position 339, and all possible combinations within CH2 and other domains.

[0214] In this embodiment, modifications may be independently and optionally selected from positions 355, 359, 362, 384, 389, 392, 397, 418, 419, 444, and 447 (EU numbering) of the CH3 domain. Specific substitutions used to reduce the pI of the CH3 domain include, but are not limited to, non-native glutamine or glutamate at position 355, non-native serine at position 384, non-native asparagine or glutamate at position 392, non-native methionine at position 397, non-native glutamate at position 419, non-native glutamate at position 359, non-native glutamate at position 362, non-native glutamate at position 389, non-native glutamate at position 418, non-native glutamate at position 444, and deletion or non-native aspartate at position 447.

[0215] In some embodiments, anti-CD28 × anti-TROP2 antibodies include an amino acid substitution in one of their Fc domains that reduces binding to protein A. Such purified variants produce heterodimers with asymmetric binding to protein A, which can be used to separate the heterodimers from the homodimer population by a pH gradient. Exemplary purified amino acid substitutions that reduce binding to protein A include, but are not limited to, H435R and Y436F (IgG1 CH3 domain, EU numbering). See, for example, US2010331527 (which is incorporated by reference in whole and in particular for relevant disclosures regarding Fc domain modifications for reducing protein A binding).

[0216] 3. Isomorphic Variants Furthermore, many embodiments of the target heterodimer antibodies rely on the “introduction” of pI amino acids at specific positions from one IgG isotype to another, thus reducing or eliminating the possibility of unwanted immunogenicity being introduced into the variant. Numerous of these are shown in Figure 21 of U.S. Publication 2014 / 0370013 (incorporated herein by reference). That is, IgG1 is a common isotype for therapeutic antibodies for a variety of reasons, including high effector function. However, the polyconstant region of IgG1 has a higher pI than that of IgG2 (8.10 vs. 7.31). By introducing IgG2 residues at specific positions into the IgG1 backbone, the pI of the resulting monomer is reduced (or increased), additionally exhibiting a longer serum half-life. For example, IgG1 has glycine at position 137 (pI 5.97), while IgG2 has glutamic acid (pI 3.22); introducing glutamic acid would affect the pI of the resulting protein. As described below, numerous amino acid substitutions are usually required to significantly affect the pI of variant antibodies. However, it should be noted that even changes in the IgG2 molecule can lead to an increase in serum half-life, as discussed below.

[0217] In other embodiments, as described further below, non-isotypic amino acid changes are made to reduce the overall charge state of the resulting protein (for example, by changing higher pI amino acids to lower pI amino acids) or to allow structural modifications for stability or other reasons.

[0218] Furthermore, significant changes can be observed in each monomer of the heterodimer by manipulating the pI of both the heavy and light steady domains. As discussed herein, by making the pI of the two monomers differ by at least 0.5, separation by ion-exchange chromatography, isoelectric focusing, or other isoelectric-sensitive methods may be possible.

[0219] 4. Calculation of pI The pI of each monomer of the antibodies provided herein may depend on the pI of the variant heavy chain constant domain as well as the pI of the total monomer including the variant heavy chain constant domain and its fusion partner. Therefore, in some embodiments, the pI change is calculated based on the variant heavy chain constant domain using the chart in Figure 19 of U.S. Publication 2014 / 0370013. As discussed herein, which monomer to manipulate is typically determined by the Fv and the pI inherent in the skeletal region. Alternatively, the pI of each monomer may be compared.

[0220] 5. pI variants that also confer better FcRn in vivo binding. If a pI variant reduces monomeric pI, the pI variant may have the added benefit of improving serum retention in vivo.

[0221] Although still under investigation, the Fc region is thought to have a longer half-life in vivo because binding to FcRn at pH 6 in endosomes sequesters Fc (Ghetie and Ward, 1997 Immunol Today. 18(12):592-598 (integrated by reference)). The endosomal compartment then recycles Fc to the cell surface. When the compartment opens to the extracellular space, a higher pH of approximately 7.4 induces the release of Fc back into the bloodstream. In mice, Dall' Acqua et al. showed that Fc mutants with increased FcRn binding at pH 6 and pH 7.4 actually had lower serum concentrations and the same half-life as wild-type Fc (Dall' Acqua et al. 2002, J.Immunol. 169:5171-5180 (integrated by reference)). The increased affinity of Fc for FcRn at pH 7.4 is thought to hinder the release and return of Fc into the bloodstream. Therefore, Fc mutations that would increase the half-life of Fc in vivo would ideally increase FcRn binding at lower pH levels while still allowing Fc release at higher pH levels. The amino acid histidine changes its charge state in the pH range of 6.0–7.4. Therefore, it is not surprising to find His residues at important positions in the Fc / FcRn complex.

[0222] Recently, it has been suggested that antibodies with variable regions having lower isoelectric points may also have longer serum half-lives (Igawa et al., 2010 PEDS.23(5):385-392 (incorporated by reference)). However, the mechanism is still not well understood. Moreover, the variable region differs from antibody to antibody. Constant region variants with reduced pI and extended half-lives would offer a more modular approach to improving the pharmacokinetic properties of the antibodies described herein.

[0223] E. Additional Fc variants for additional functionality In addition to the heterodimerization variants discussed above, there are numerous useful Fc amino acid modifications that can be performed for a variety of reasons, including, but not limited to, altering binding to one or more FcγR receptors and modified binding to FcRn receptors, as discussed below.

[0224] Accordingly, the antibodies (heterodimers and homodimers) provided herein may include such amino acid modifications with or without the heterodimerizing variants (e.g., pI variants and stereovariants) outlined herein. Each set of variants may be independently and optionally included in or excluded from any particular heterodimer protein.

[0225] 1. FcγR and FcRn variants Therefore, there are numerous useful Fc substitutions that can be performed to alter the binding of one or more FcγR receptors. In a given embodiment, the target antibody contains modifications (i.e., "FcγR variants") that alter the binding of one or more FcγR receptors. Substitutions resulting in increased and decreased binding may be useful. For example, increased binding to FcγRIIIa is known to typically result in increased ADCC (antibody-dependent cell-mediated cytotoxicity; a cell-mediated reaction in which nonspecific cytotoxic cells expressing FcγR recognize the bound antibody on target cells, subsequently causing lysis of the target cells). Similarly, decreased binding to FcγRIIb (an inhibitory receptor) may be equally beneficial in several situations. The amino acid substitutions used in the target antibodies include those listed in U.S. Patent Nos. 8,188,321 (particularly Figure 41) and 8,084,582, and U.S. Publication Nos. 20060235208 and 20070148170 (all of which are expressly incorporated herein by reference with respect to the variants disclosed therein that affect them in whole and in particular to Fcγ receptor binding). Specific variants used include, but are not limited to, 236A, 239D, 239E, 332E, 332D, 239D / 332E, 267D, 267E, 328F, 267E / 328F, 236A / 332E, 239D / 332E / 330Y, 239D, 332E / 330L, 243A, 243L, 264A, 264V, and 299T. Such modifications may be present in one or both Fc domains of the target antibody.

[0226] In some embodiments, the target antibody includes one or more Fc modifications that increase the serum half-life. These include, but are not limited to, 434S, 434A, 428L, 308F, 259I, 428L / 434S, 259I / 308F, 436I / 428L, 436I or V / 434S, 436V / 428L, 259I / 308F / 428L, and M252Y / S254T / T256E, as specifically disclosed in USSN12 / 341,769 (the entirety of which is incorporated herein by reference), Fc substitutions utilized for increased binding to the FcRn receptor and increased serum half-life. Such modifications may be contained in one or both Fc domains of the target antibody.

[0227] 2. Decline variant In some embodiments, heterodimer antibodies include one or more modifications that reduce or eliminate the normal binding of the Fc domain to one or more or all of the Fcγ receptors (e.g., FcγR1, FcγRIIa, FcγRIIb, FcγRIIIa, etc.) in order to avoid additional mechanisms of action. Such modifications are referred to as “FcγR reduction variants” or “Fc knockout (FcKO or KO)” variants. In these embodiments, for some therapeutic applications, it is desirable to reduce or eliminate the normal binding of the Fc domain to one or more or all of the Fcγ receptors (e.g., FcγR1, FcγRIIa, FcγRIIb, FcγRIIIa, etc.) in order to avoid additional mechanisms of action. That is, for example, in many embodiments, particularly in the use of bispecific antibodies that monovalently bind CD28, it is usually desirable to reduce FcγRIIIa binding in order to eliminate or significantly reduce ADCC activity. In some embodiments of the target antibody described herein, at least one of the Fc domains contains one or more Fcγ receptor attenuation variants. In some embodiments of the target antibody described herein, both of the Fc domains contain one or more Fcγ receptor attenuation variants. These attenuation variants are shown in Figure 5, each of which may be included or excluded independently and optionally. A preferred embodiment utilizes attenuation variants selected from the group consisting of L234A / L235A / D265S, G236R / L328R, E233P / L234V / L235A / G236del / S239K, E233P / L234V / L235A / G236del / S267K, E233P / L234V / L235A / G236del / S239K / A327G, E233P / L234V / L235A / G236del / S267K / A327G, and E233P / L234V / L235A / G236del. It should be noted that the diminishing variants referred to herein diminish FcγR binding but do not typically diminish FcRn binding.

[0228] As is known in the art, the Fc domain of human IgG1 has the highest binding to the Fcγ receptor, and therefore the reduced-binding variant can be used when the constant domain (or Fc domain) in the heterodimeric antibody backbone is IgG1. Alternatives to, or in addition to, the reduced-binding variant in the IgG1 background, mutations at glycosylation site 297 (usually to A or S) can significantly reduce binding to FcγRIIIa, for example. Human IgG2 and IgG4 have naturally reduced binding to the Fcγ receptor, and therefore their backbones can be used with or without the reduced-binding variant.

[0229] F. Combination of heterodimer and Fc variant As will be understood by those skilled in the art, all of the listed heterodimerized variants (including skewed and / or purified variants) can be combined in any way, independently and arbitrarily, as long as they retain their "strandiness" or "monomer partitioning." Furthermore, all of these variants can be combined with any of the heterodimerized forms.

[0230] While the most commonly used embodiments for the pI variant are shown in the figure, other combinations can be generated according to fundamental rules for modifying the pI difference between the two monomers to facilitate purification.

[0231] Furthermore, both heterodimerized variants (skewed and purified variants) can be independently and optionally combined with Fc-decaying variants, Fc variants, and FcRn variants, as generally outlined herein.

[0232] Exemplary combinations of variants included in some embodiments of heterodimer 1+1 Fab-scFv-Fc, 2+1 mAb-Fc, and 2+1 Fab2-scFv-Fc form antibodies are shown in Figure 8. In some embodiments, the heterodimer antibodies include the variant combinations shown in Figure 8.

[0233] G. Useful antibody formats As will be understood by those skilled in the art and will be discussed more thoroughly below, the heterodimer bispecific antibodies provided herein can take on several different configurations, as schematically shown in Figure 15.

[0234] As will be understood by those skilled in the art, the heterodimer form of the present invention may have different valencies and may be bispecific. That is, the heterodimer antibody of the present invention may be bivalent and bispecific, or trivalent and bispecific, where the first antigen is bound by two binding domains and the second antigen is bound by a second binding domain. As outlined herein, when CD28 is one of the target antigens, it is preferable that CD28 binds only in a monovalent state.

[0235] The present invention utilizes a CD28 binding domain in combination with a TROP2 binding domain. As will be understood by those skilled in the art, any combination of anti-CD28 CDRs, anti-CD28 variable light domains and variable heavy domains, Fab and scFv or their variants shown in any of the figures (see in particular Figures 16-19 and 22) may be used. Similarly, regardless of which of the CDRs, variable light domains and variable heavy domains, Fab and scFv or their variants shown in any of the figures (e.g., Figures 23-28) may be used, any of the TROP2 antigen-binding domains may be used and may be combined in any combination, either independently or arbitrarily.

[0236] 1.1+1 CLC format One heterodimer antibody form particularly utilized in the anti-CD28 × anti-TROP2 antibodies provided herein is the "1+1 common light chain" or "1+1 CLC" form, shown in Figure 15C. The 1+1 CLC form antibody comprises a first monomer containing VH1-CH1-hinge-CH2-CH3 (where VH1 is the first variable weight domain and CH2-CH3 is the first Fc domain), a second monomer containing VH2-CH1-hinge-CH2-CH3 (where VH2 is the second variable weight domain and CH2-CH3 is the second Fc domain), and a third monomer "common light chain" containing VL-CL (where VL is the common variable light domain and CL is the constant light domain). In some embodiments, the 1+1 CLC contains two identical common light chains. In such embodiments, one VL of the common light chain pairs with VH1 to form a first binding domain (i.e., CD28 or TROP2) having a first antigen-binding specificity, and the other VL of the common light chain pairs with VH2 to form a second binding domain (i.e., CD28 or TROP2) having a second antigen-binding specificity. In some embodiments, the 1+1 CLC-type antibody is a bivalent antibody.

[0237] In some embodiments, the first and second Fc domains of the 1+1 CLC form are variant Fc domains containing heterodimerized scuba riants (e.g., the set of amino acid substitutions shown in Figures 3 and 9). Particularly useful heterodimerized scuba riants include S364K / E357Q:L368D / K370S;L368D / K370S:S364K;L368E / K370S:S364K;T411T / E360E / Q362E:D401K;L368D / K370S:S364K / E357L;K370S:S364K / E357Q;T366S / L368A / Y407V:T366W and T366S / L368A / Y407V / Y349C:T366W / S354C (EU numbering)). In an exemplary embodiment, one of the first or second variant Fc domains contains the heterodimerized scuba riant L368D / K370S, and the other of the first or second variant Fc domains contains the heterodimerized scuba riant S364K / E357Q, with numbering following EU numbering.

[0238] In some embodiments, the variant Fc domain includes a de-evolution variant (including the one shown in Figure 5). In some embodiments, each of the first and second variant Fc domains includes a de-evolution variant E233P / L234V / L235A / G236_ / S267K, numbered according to EU numbering.

[0239] In some embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first or second monomer includes a pI variant (including those shown in Figure 4). In exemplary embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first or second monomer includes pI variants N208D / Q295E / N384D / Q418E / N421D, numbered according to EU numbering.

[0240] In some embodiments, the 1+1 CLC-type antibodies provided herein include the FcRn variant M428L / N434S, which is numbered according to EU numbering.

[0241] In exemplary embodiments, the first variant Fc domain contains the heterodimerized scuba rianto S364K / E357Q, the second variant Fc domain contains the heterodimerized scuba rianto L368D / K370S, each of the first and second variant Fc domains contains the diminished variant E233P / L234V / L235A / G236_ / S267K, and the steady-state domain (CH1-hinge-CH2-CH3) of the second monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, with numbering following EU numbering.

[0242] In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant L368D / K370S / N208D / Q295E / N384D / Q418E / N421D / E233P / L234V / L235A / G236del / S267K, and the first Fc domain contains the amino acid variant S364K / E357Q / E233P / L234V / L235A / G236del / S267K, with the numbering following EU numbering.

[0243] In some embodiments, the 1+1 CLC-type antibody provided herein further comprises the FcRn variant M428L / N434S, the numbering of which follows EU numbering.

[0244] In some embodiments, one of the first or second binding domains binds to CD28, and the other binding domain binds to TROP2. The 1+1 CLC-type antibody of interest may contain any preferred CD28-binding domain and TROP2-binding domain, including either the CD28-binding domain and the TROP2-binding domain or a variant thereof, as provided herein.

[0245] In some embodiments, the CD28 binding domain includes vhCDR1-3 of one of the CD28 variable heavy domains shown in Figure 34 and vlCDR1-3 of the 1F11-1A3.315 L1 variable light domain (see Figure 36A). In some embodiments, the CD28 binding domain includes vhCDR1-3 of one of the following CD28 variable heavy domains: 1A7[CD28]_H1.129 and 1A7[CD28]_H1.106 (Figure 34) and vlCDR1-3 of the 1F11-1A3.315 L1 variable light domain (see Figure 36A). In another exemplary embodiment, the CD28 ABD of the target anti-CD28 × anti-TROP2 antibody comprises one variable weight (VH) domain or a variant thereof from the CD28 variable weight domains shown in Figure 34, and the 1F11-1A3.315 L1 variable light domain (see Figure 36A) or a variant thereof. In some embodiments, the CD28 binding domain comprises one variable weight domain or a variant thereof from the following CD28 variable weight domains: 1A7[CD28]_H1.129 and 1A7[CD28]_H1.106 (Figure 34), and the 1F11-1A3.315 L1 variable light domain (see Figure 36A) or a variant thereof.

[0246] In some embodiments, the TROP2 binding domain includes vhCDR1-3 of one of the TROP2 variable heavy domains shown in Figures 36 and 52-56, and vlCDR1-3 of the 1F11-1A3.315 L1 variable light domain (see Figure 36A). In some embodiments, the TROP2 binding domain includes vhCDR1-3 of one of the following TROP2 variable heavy domains: 1F11-1A3.315[Trop2]_H1, 1F11-1A3.315[Trop2]_H1.10, 1F11-1A3.315[Trop2]_H1.19, 2C5A3.316[TROP2]_H1, 1A3A4.312[TROP2]_H1, 1B2A4.312[TROP2]_H1, 1B11A3.316[TROP2]_H1, and 1C9A4.313[TROP2]_H1 (Figures 36 and 52-56), and vlCDR1-3 of the 1F11-1A3.315 L1 variable light domain (see Figure 36A). In another exemplary embodiment, the TROP2-binding domain of the anti-CD28×anti-TROP2 antibody in question includes one variable weight (VH) domain or a variant thereof from the TROP2 variable weight domains shown in Figure 36, and the 1F11-1A3.315 L1 variable light domain (see Figure 36A) or a variant thereof. In some embodiments, the TROP2 binding domain includes one of the following TROP2 variable heavy domains: 1F11-1A3.315[Trop2]_H1, 1F11-1A3.315[Trop2]_H1.10, 1F11-1A3.315[Trop2]_H1.19, 2C5A3.316[TROP2]_H1, 1A3A4.312[TROP2]_H1, 1B2A4.312[TROP2]_H1, 1B11A3.316[TROP2]_H1, and 1C9A4.313[TROP2]_H1 (Figures 36 and 52-56), or a variant thereof, and the 1F11-1A3.315 L1 variable light domain (see Figure 36A), or a variant thereof.

[0247] An exemplary target anti-CD28 × anti-TROP2 antibody in 1+1 CLC format is shown in Figure 38.

[0248] 2.2+1 CLC format Another heterodimer antibody form particularly utilized in the CD28× anti-TROP2 antibodies provided herein is the “2+1 common light chain” or “2+1 CLC” form, shown in Figure 15D. The 2+1 CLC form comprises a first monomer containing VH1-CH1-linker-VH1-CH1-hinge-CH2-CH3 (where VH1 is the first variable weight domain and CH2-CH3 is the first Fc domain), a second monomer containing VH2-CH1-hinge-CH2-CH3 (where VH2 is the second variable weight domain and CH2-CH3 is the second Fc domain), and a third monomer containing the “common light chain” VL-CL (where VL is the common variable light domain and CL is the constant light domain). In some embodiments, the 2+1 CLC form comprises three identical common light chains (i.e., the first, second, and third common light chains). In such embodiments, the VLs of the first and second common light chains each pair with one of the VH1s of the first monomer to form two first binding domains having a first antigen-binding specificity, and the VL of the third common light chain pairs with VH2 to form a second binding domain having a second antigen-binding specificity. The linker of the first monomer can be any preferred linker, including any one or a combination thereof of the domain linkers shown in Figure 7. In some embodiments, the 2+1 CLC-type antibody is a trivalent antibody.

[0249] In some embodiments, the second and second Fc domains of the 2+1 CLC form are variant Fc domains containing heterodimerized scuba riants (e.g., the set of amino acid substitutions shown in Figures 3 and 9). Particularly useful heterodimerized scuba riants include S364K / E357Q:L368D / K370S;L368D / K370S:S364K;L368E / K370S:S364K;T411T / E360E / Q362E:D401K;L368D / K370S:S364K / E357L;K370S:S364K / E357Q;T366S / L368A / Y407V:T366W and T366S / L368A / Y407V / Y349C:T366W / S354C (EU numbering)). In an exemplary embodiment, one of the first or second variant Fc domains contains the heterodimerized scuba riant L368D / K370S, and the other of the first or second variant Fc domains contains the heterodimerized scuba riant S364K / E357Q, with numbering following EU numbering.

[0250] In some embodiments, the variant Fc domain includes a de-evolution variant (including the one shown in Figure 5). In some embodiments, each of the first and second variant Fc domains includes a de-evolution variant E233P / L234V / L235A / G236_ / S267K, numbered according to EU numbering.

[0251] In some embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first or second monomer includes a pI variant (including those shown in Figure 4). In exemplary embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first or second monomer includes pI variants N208D / Q295E / N384D / Q418E / N421D, numbered according to EU numbering.

[0252] In some embodiments, the 2+1 CLC-type antibodies provided herein further include the FcRn variant M428L / N434S, which is numbered according to EU numbering.

[0253] In exemplary embodiments, the first variant Fc domain contains the heterodimerized scuba rianto L368D / K370S, the second variant Fc domain contains the heterodimerized scuba rianto S364K / E357Q, each of the first and second variant Fc domains contains the diminished variant E233P / L234V / L235A / G236_ / S267K, and the constant domain (CH1-hinge-CH2-CH3) of the first monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, numbered according to EU numbering. In some embodiments, the 2+1 CLC-type antibody provided herein further contains the FcRn variant M428L / N434S, numbered according to EU numbering.

[0254] In some embodiments, the CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant L368D / K370S / N208D / Q295E / N384D / Q418E / N421D / E233P / L234V / L235A / G236del / S267K, and the first Fc domain contains the amino acid variant S364K / E357Q / E233P / L234V / L235A / G236del / S267K, with the numbering following EU numbering.

[0255] In some embodiments, each of the two first binding domains binds to tumor TROP2, and the second binding domain binds to CD28. The 2+1 CLC-type antibody of interest may contain any preferred CD28-binding domain and TROP2 domain, including either or a variant thereof of the CD28-binding domain and TROP2-binding domain provided herein.

[0256] In some embodiments, the CD28 binding domain includes vhCDR1-3 of one of the CD28 variable heavy domains shown in Figure 34 and vlCDR1-3 of the 1F11-1A3.315 L1 variable light domain (see Figure 36A). In some embodiments, the CD28 binding domain includes vhCDR1-3 of one of the following CD28 variable heavy domains: 1A7[CD28]_H1.129 and 1A7[CD28]_H1.106 (Figure 34) and vlCDR1-3 of the 1F11-1A3.315 L1 variable light domain (see Figure 36A). In another exemplary embodiment, the CD28 ABD of the target anti-CD28 × anti-TROP2 antibody comprises one variable weight (VH) domain or a variant thereof from the CD28 variable weight domains shown in Figure 34, and the 1F11-1A3.315 L1 variable light domain (see Figure 36A) or a variant thereof. In some embodiments, the CD28 binding domain comprises one variable weight domain or a variant thereof from the following CD28 variable weight domains: 1A7[CD28]_H1.129 and 1A7[CD28]_H1.106 (Figure 34), and the 1F11-1A3.315 L1 variable light domain (see Figure 36A) or a variant thereof.

[0257] In some embodiments, the TROP2 binding domain includes vhCDR1-3 of one of the TROP2 variable heavy domains shown in Figures 36 and 52-56, and vlCDR1-3 of the 1F11-1A3.315 L1 variable light domain (see Figure 36A). In some embodiments, each TROP2-binding domain includes vhCDR1-3 of one of the following TROP2 variable heavy domains: 1F11-1A3.315[Trop2]_H1, and 1F11-1A3.315[Trop2]_H1.19, 2C5A3.316[TROP2]_H1, 1A3A4.312[TROP2]_H1, 1B2A4.312[TROP2]_H1, 1B11A3.316[TROP2]_H1, and 1C9A4.313[TROP2]_H1 (Figures 36 and 52-56), and vlCDR1-3 of the 1F11-1A3.315 L1 variable light domain (see Figure 36A). In another exemplary embodiment, the TROP2-binding domain of the anti-CD28×anti-TROP2 antibody in question comprises one variable weight (VH) domain or a variant thereof from the TROP2 variable weight domains shown in Figures 36 and 52-56, and the 1F11-1A3.315 L1 variable light domain (see Figure 36A) or a variant thereof. In some embodiments, each TROP2-binding domain includes one of the following TROP2 variable heavy domains: 1F11-1A3.315[Trop2]_H1, and a variant thereof from among 1F11-1A3.315[Trop2]_H1.19, 2C5A3.316[TROP2]_H1, 1A3A4.312[TROP2]_H1, 1B2A4.312[TROP2]_H1, 1B11A3.316[TROP2]_H1, and 1C9A4.313[TROP2]_H1 (Figures 36 and 52-56), and a variant thereof from among 1F11-1A3.315 L1 variable light domain (see Figure 36A) or a variant thereof.

[0258] An exemplary target anti-CD28 × anti-TROP2 antibody in 2+1 CLC format is shown in Figure 39.

[0259] 3.1+1 Fab-scFv-Fc format One heterodimer antibody form particularly utilized in the anti-CD28 × anti-TROP2 antibodies provided herein is the “1+1 Fab-scFv-Fc” or “bottle opener” form, as shown in Figure 15A. The 1+1 Fab-scFv-Fc form antibody contains a first monomer which is a “standard” heavy chain (VH1-CH1-hinge-CH2-CH3), where VH1 is the first variable heavy domain and CH2-CH3 is the first Fc domain. The 1+1 Fab-scFv-Fc also contains a light chain which includes a first variable light domain VL1 and a constant light domain CL. The light chain interacts with VH1-CH1 of the first monomer to form a first antigen-binding domain which is Fab. The second monomer of the antibody contains a second binding domain which is a single-stranded Fv (defined hereafter as “scFv”) and a second Fc domain. The scFv comprises a second variable heavy domain (VH2) and a second variable light domain (VL2), where VH2 is bound to VL2 using a potentially charged scFv linker (see, e.g., Figure 6). The scFv is bound to the heavy chain using a domain linker (see, e.g., Figure 7). The two monomers are brought about by the use of amino acid variants (e.g., heterodimerizing variants discussed above) in constant regions (e.g., Fc domain, CH1 domain, and / or hinge region) that promote the formation of a heterodimer antibody, as will be described more thoroughly below. This structure is roughly visually similar to a bottle opener and is therefore sometimes referred to herein as the “bottle opener” form. In some embodiments, the 1+1 Fab-scFv-Fc form antibody is a bivalent antibody.

[0260] The "1+1 Fab-scFv-Fc" format offers several unique advantages. As is known in the art, antibody analogs that rely on two scFv constructs often have stability and aggregation problems, which can be mitigated in this invention by the addition of "standard" heavy-chain and light-chain pairings. Furthermore, in contrast to formats that rely on two heavy chains and two light chains, there is no problem with incorrect pairing of heavy and light chains (e.g., heavy 1 pairing with light 2).

[0261] In some embodiments of the 1+1 Fab-scFv-Fc antibody, one of the first or second antigen-binding domains is a CD28-binding domain, and the other binding domain is a TROP2-binding domain. In some embodiments of 1+1 Fab-scFv-Fc, it is an scFv that binds to CD28 and a Fab that binds to TROP2. An exemplary anti-CD28 × anti-TROP2 bispecific antibody of the 1+1 Fab-scFv-Fc form is shown in Figure 30.

[0262] In some embodiments, the first and second Fc domains of a 1+1 Fab-scFv-Fc antibody are variant Fc domains containing heterodimerized scuba riants (e.g., the set of amino acid substitutions shown in Figures 3 and 9). Particularly useful heterodimerized scuba riants include S364K / E357Q:L368D / K370S;L368D / K370S:S364K;L368E / K370S:S364K;T411T / E360E / Q362E:D401K;L368D / K370S:S364K / E357L;K370S:S364K / E357Q;T366S / L368A / Y407V:T366W and T366S / L368A / Y407V / Y349C:T366W / S354C (EU numbering)). In an exemplary embodiment, one of the first or second variant Fc domains contains the heterodimerized scuba riant L368D / K370S, and the other of the first or second variant Fc domains contains the heterodimerized scuba riant S364K / E357Q, with numbering following EU numbering.

[0263] In some embodiments, the variant Fc domain includes a de-evolution variant (including the one shown in Figure 5). In some embodiments, each of the first and second variant Fc domains includes a de-evolution variant E233P / L234V / L235A / G236_ / S267K, numbered according to EU numbering.

[0264] In some embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first monomer includes pI variants (including those shown in Figures 3 and 4). In exemplary embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first monomer includes pI variants N208D / Q295E / N384D / Q418E / N421D, numbered according to EU numbering.

[0265] In exemplary embodiments, the 1+1 Fab-scFv-Fc form antibody contains the amino acid modification combination shown in Figure 8. In such embodiments, the CH1-hinge-CH2-CH3 of the first monomer contains the amino acid variant L368D / K370S / N208D / Q295E / N384D / Q418E / N421D / E233P / L234V / L235A / G236del / S267K, and the second Fc domain contains the amino acid variant S364K / E357Q / E233P / L234V / L235A / G236del / S267K, with the numbering following EU numbering.

[0266] In some embodiments, the scFv of the 1+1 Fab-scFv-Fc antibody provided herein includes a charged scFv linker (including the one shown in Figure 6). In some embodiments, the 1+1 Fab-scFv-Fc antibody provided herein includes the FcRn variant M428L / N434S, which is numbered according to EU numbering.

[0267] In the exemplary embodiment of the 1+1 Fab-scFv-Fc antibody, the first Fc domain contains the heterodimerized scuba riant L368D / K370S, the second Fc domain contains the heterodimerized scuba riant S364K / E357Q, each of the first and second Fc domains contains the diminished variant E233P / L234V / L235A / G236_ / S267K, and the constant domain (CH1-hinge-CH2-CH3) of the first monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, with numbering following EU numbering. In some embodiments, the scFv of the 1+1 Fab-scFv-Fc antibody provided herein contains the (GKPGS) 4-charged scFv linker (SEQ ID NO: 24). In some embodiments, the 1+1 Fab-scFv-Fc format antibody provided herein includes the FcRn variant M428L / N434S, and the numbering follows EU numbering. In some embodiments, the scFv of the 1+1 Fab-scFv-Fc format antibody provided herein includes a charged scFv linker (including one shown in Figure 6).

[0268] The target 1+1 Fab-scFv-Fc antibody may contain any suitable CD28-binding domain, including any of the CD28-binding domains provided herein. In some embodiments, the CD28-binding domain is one of the following CD28-binding domains or variants thereof: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1.71, CD28.3[CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28]_H 0L0, TGN1412_H1L1, 341VL34[CD28]_H1L1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9.3[CD28]_H0L0, hu9.3[CD28]_H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, and TN228[CD28]_H4L2 (Figures 16, 19, and 22). Exemplary additional VH and VL sequences of the CD28-binding domain that can be used in the target 1+1 Fab-scFv-Fc type antibody are shown in Figures 16, 17, 18, 34, and 36.

[0269] In some embodiments of the 1+1 Fab-scFv-Fc format, the anti-CD28 ABD has VH and VL domains selected from the following:

[0270] (i) VHs containing vhCDR1, vhCDR2, and vhCDR3, each having the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of any VH or variant thereof of the CD28 binding domain shown in Figures 16, 19, and 22; and (ii) VLs containing vlCDR1, vlCDR2, and vlCDR3, each having the amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of any VL or variant thereof of the CD28 binding domain shown in Figures 16, 19, and 22;

[0271] (i) a VH containing vhCDR1, vhCDR2, and vhCDR3, having the respective amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of the VH or its variant shown in Figures 16, 17, or 34; and (ii) a VL containing vlCDR1, vlCDR2, and vlCDR3, having the respective amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of the VL or its variant shown in Figures 16, 18, or 36; or

[0272] (i) VH containing vhCDR1, vhCDR2, and vhCDR3, each having the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of the VH or its variant shown in Figure 34; and (ii) VL containing vlCDR1, vlCDR2, and vlCDR3, each having the amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of the 1F11-1A3.315L1 variable light domain (see Figure 36A) or its variant.

[0273] In some embodiments of the 1+1 Fab-scFv-Fc format, the anti-CD28 ABD has VH and VL domains selected from the following:

[0274] (i) VH having the amino acid sequence of any VH or a variant thereof of the CD28 binding domain shown in Figures 16, 19, and 22; and (ii) VL having the amino acid sequence of any VL or a variant thereof of the CD28 binding domain shown in Figures 16, 19, and 22;

[0275] (i) VH having the amino acid sequence of VH or its variant shown in Figure 16, 17, or 34, and (ii) VL having the amino acid sequence of VL or its variant shown in Figure 16, 18, or 36; or

[0276] (i) VH having the amino acid sequence of VH or its variant shown in Figure 34, and (ii) VL having the amino acid sequence of the 1F11-1A3.315 L1 variable light domain (see Figure 36A) or its variant.

[0277] The target 1+1 Fab-scFv-Fc antibody may contain any suitable TROP2-binding domain, including any of the TROP2-binding domains provided herein. Exemplary TROP2-binding domains that can be used in the target 1+1 Fab-scFv-Fc antibody are shown in Figure 23. Additional VH and VL sequences of exemplary TROP2-binding domains that can be used in the target 1+1 Fab-scFv-Fc antibody are shown in Figures 24-28, 36, and 52-56.

[0278] In some embodiments of the 1+1 Fab-scFv-Fc form, the anti-TROP2 ABD has VH and VL domains selected from the following:

[0279] (i) a VH having the amino acid sequence of any one of the TROP2 binding domains shown in Figure 23 or a variant thereof, and (ii) a VL having the amino acid sequence of the TROP2 binding domain or a variant thereof shown in Figure 23;

[0280] (i) VH having the amino acid sequence of VH or its variant shown in Figures 24-28, and (ii) VL having the amino acid sequence of VL or its variant shown in Figures 24-28; or

[0281] (i) VH having the amino acid sequence of VH or its variant shown in Figure 36, and (ii) VL having the amino acid sequence of the 1F11-1A3.315 L1 variable light domain (see Figure 36A) or its variant.

[0282] Figure 10 shows several exemplary Fc domain sequences useful in 1+1 Fab-scFv-Fc type antibodies. The “monomer 1” sequence shown in Figure 10 typically refers to the Fc domain of the “Fab-Fc heavy chain,” and the “monomer 2” sequence refers to the Fc domain of the “scFv-Fc heavy chain.” Figures 12 and 13 also provide exemplary CH1-hinge domains, CH1 domains, and hinge domains that may be included in the first or second monomer of the 1+1 Fab-scFv-Fc form. Furthermore, Figure 14 provides useful CL sequences that can be used in this form.

[0283] 4.2+1 mAb-scFv format One heterodimer antibody form particularly used in the target bispecific anti-CD28 × anti-TROP2 antibody is the 2+1 mAb-scFv form shown in Figure 15E. This antibody form contains three antigen-binding domains, namely two Fab moieties and an scFv bound to one C-terminus of the heavy chain. In some embodiments of this form, each Fab moiety binds to TROP2 (in this case, human TROP2), and the "additional" scFv domain binds to CD28. In other words, this mAb-scFv form is a trivalent antibody.

[0284] In these embodiments, the first chain or monomer comprises VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, and the second monomer comprises VH1-CH1-hinge-CH2-CH3-domain linker-scFv domain from the N-terminus to the C-terminus, where the scFv domain comprises a second VH(VH2), a second VL(VL2), and the scFv linker. The antibody also comprises two light chains comprising VL1-CL, respectively, where VL1 is the first variable light domain. The scFv domain comprises a second VH(VH2), a second VL(VL2), and the scFv linker. In some embodiments, VH1 of the first monomer and VH1 of one of the two light chains, and VH1 of the second monomer and VL1 of the other of the two light chains, respectively, form a first antigen-binding domain (ABD). VH2 and VL2 form a second ABD. In some embodiments, the first ABD binds to human TROP2, and the second ABD binds to human CD28.

[0285] For all scFv domains as described herein, the scFv domain may be oriented from the N-terminus to the C-terminus in either the VH2-scFv linker-VL2 or VL2-scFv linker-VH2 orientation. Thus, the second monomer may contain VH1-CH1-hinge-CH2-CH3-domain linker-VH2-scFv linker-VL2 or VH1-CH1-hinge-CH2-CH3-domain linker-VL2-scFv linker-VH2.

[0286] In some embodiments, the first and second Fc domains of a 2+1 mAb-scFv antibody are variant Fc domains containing heterodimerized scuba riants (e.g., the set of amino acid substitutions shown in Figures 3 and 9). Particularly useful heterodimerized scuba riants include S364K / E357Q:L368D / K370S;L368D / K370S:S364K;L368E / K370S:S364K;T411T / E360E / Q362E:D401K;L368D / K370S:S364K / E357L;K370S:S364K / E357Q;T366S / L368A / Y407V:T366W and T366S / L368A / Y407V / Y349C:T366W / S354C (EU numbering)). In an exemplary embodiment, one of the first or second variant Fc domains contains the heterodimerized scuba riant L368D / K370S, and the other of the first or second variant Fc domains contains the heterodimerized scuba riant S364K / E357Q, with numbering following EU numbering.

[0287] In some embodiments, the variant Fc domain includes a de-evolution variant (including the one shown in Figure 5). In some embodiments, each of the first and second variant Fc domains includes a de-evolution variant E233P / L234V / L235A / G236_ / S267K, numbered according to EU numbering.

[0288] In some embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first monomer includes a pI variant (including those shown in Figure 4). In exemplary embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first monomer includes pI variants N208D / Q295E / N384D / Q418E / N421D, numbered according to EU numbering.

[0289] In some embodiments, the scFv of the 2+1 mAb-scFv antibody provided herein includes a charged scFv linker (including the one shown in Figure 6). In some embodiments, the 2+1 mAb-scFv antibody provided herein includes the FcRn variant M428L / N434S, which is numbered according to EU numbering.

[0290] In exemplary embodiments, the 2+1 mAb-scFv form antibody includes the amino acid modification combinations shown in Figure 8. In such embodiments, the first variant Fc domain contains the heterodimerized scuba rianto L368D / K370S, the second variant Fc domain contains the heterodimerized scuba rianto S364K / E357Q, each of the first and second variant Fc domains contains the diminished variant E233P / L234V / L235A / G236_ / S267K, and the constant domain (CH1-hinge-CH2-CH3) of the first monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, with numbering following EU numbering. In some embodiments, the scFv of the 2+1 mAb-scFv antibody provided herein includes a (GKPGS) 4-charged scFv linker (SEQ ID NO: 24). In some embodiments, the 2+1 mAb-scFv antibody provided herein includes the FcRn variant M428L / N434S, which is numbered according to EU numbering.

[0291] In some embodiments, the scFv of the second monomer of a 2+1 mAb-scFv antibody is CD28-conjugated, and the VH1 of the first and second monomers and the VL1 of the common light chain each form a TROP2-binding domain. The 2+1 mAb-scFv antibody may include any preferred TROP2-binding domain and CD28-binding domain, including the TROP2-binding domain and CD28-binding domain provided herein, as well as any of the associated VH and VL or their variants (see, for example, Figures 16-19, 22-28, 34, 36, and 52-56).

[0292] In some embodiments of the 2+1 mAb-scFv format, the anti-CD28 ABD has VH and VL domains selected from the following:

[0293] (i) VH containing vhCDR1, vhCDR2, and vhCDR3, wherein each VH has the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of any VH or variant thereof shown in Figures 16, 19, and 22; and (ii) VL containing vlCDR1, vlCDR2, and vlCDR3, wherein each VL has the amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of any VL or variant thereof shown in Figures 16, 19, and 22; or

[0294] (i) a VH containing vhCDR1, vhCDR2, and vhCDR3, each having the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of the VH or its variant shown in Figures 16, 17, or 34; and (ii) a VL containing vlCDR1, vlCDR2, and vlCDR3, each having the amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of the VL or its variant shown in Figures 16, 18, or 36.

[0295] In some embodiments of the 2+1 mAb-scFv format, the anti-CD28 ABD has VH and VL domains selected from the following:

[0296] (i) a VH having the amino acid sequence of any VH or a variant thereof of the CD28 binding domain shown in Figures 16, 19, and 22, and (ii) a VL having the amino acid sequence of any VL or a variant thereof of the CD28 binding domain shown in Figures 16, 19, and 22; or

[0297] (i) VH having the amino acid sequence of VH or a variant thereof shown in Figure 16, 17, or 34, and (ii) VL having the amino acid sequence of VL or a variant thereof shown in Figure 16, 18, or 36.

[0298] In some embodiments of the 2+1 mAb-scFv format, each of the anti-TROP2 ABDs has VH and VL domains selected from the following:

[0299] (i) a VH having the amino acid sequence of any one of the TROP2 binding domains shown in Figure 23 or a variant thereof, and (ii) a VL having the amino acid sequence of the TROP2 binding domain or a variant thereof shown in Figure 23; or

[0300] (i) VH having the amino acid sequence of VH or its variant shown in Figures 24-28, and (ii) VL having the amino acid sequence of VL or its variant VL shown in Figures 24-28.

[0301] Figure 11 shows several exemplary Fc domain sequences useful in 2+1 mAb-scFv format antibodies. Figures 12 and 13 provide exemplary CH1-hinge domains, CH1 domains, and hinge domains that may be included in the first or second monomer of the 2+1 mAb-scFv format. Furthermore, Figure 14 provides useful CL sequences that may be used in this format.

[0302] An exemplary anti-CD28 × anti-TROP2 antibody in 2+1 mAb-scFv format is shown in Figure 31.

[0303] 5.2+1 Fab2-scFv-Fc format One heterodimeric antibody form particularly utilized in the anti-CD28 × anti-TROP2 antibodies provided herein is the 2+1 Fab2-scFv-Fc form (also referred to as the “central-scFv form”), shown in Figure 15B. This antibody form includes three antigen-binding domains: two Fab moieties and an scFv inserted between the VH-CH1 and CH2-CH3 regions of one of the monomers. In some embodiments of this form, each Fab moiety binds to TROP2, and the “additional” scFv domain binds to CD28. In some embodiments, the 2+1 Fab2-scFv-Fc form antibody is a trivalent antibody.

[0304] In some embodiments of the 2+1 Fab2-scFv-Fc form, the first monomer includes a standard heavy chain (i.e., VH1-CH1-hinge-CH2-CH3), where VH1 is the first variable weight domain and CH2-CH3 is the first Fc domain. The second monomer includes another first variable weight domain (VH1), a CH1 domain (and an optional hinge), a second Fc domain, and an scFv containing an scFv variable light domain (VL2), an scFv linker, and an scFv variable weight domain (VH2). The scFv is covalently linked between the C-terminus of the CH1 domain of the second monomer and the N-terminus of the second Fc domain using an arbitrary domain linker (VH1-CH1-[arbitrary linker]-VH2-scFv linker-VH2-[arbitrary linker]-CH2-CH3, or the opposite orientation for scFv: VH1-CH1-[arbitrary linker]-VL2-scFv linker-VH2-[arbitrary linker]-CH2-CH3). The arbitrary linker can be any suitable peptide linker, including, for example, the domain linker shown in Figure 7. This embodiment further utilizes first and second common light chains, each containing a first variable light domain (VL1) and a constant light domain (CL), respectively. The VH1-CH1 of the first and second monomers interact with one of the two common light chains, respectively, to form two identical Fabs. In some embodiments, the identical Fab is a TROP2-binding domain, and the scFv is a CD28-binding domain. As with many of the embodiments described herein, these constructs may include, as desired and as described herein, scuba riants, pI variants, attenuation variants, additional Fc variants, and the like.

[0305] In some embodiments, the second and second Fc domains of the 2+1 Fab2-scFv-Fc antibody are variant Fc domains containing heterodimerized scuba riants (e.g., the set of amino acid substitutions shown in Figures 3 and 9). Particularly useful heterodimerized scuba riants include S364K / E357Q:L368D / K370S;L368D / K370S:S364K;L368E / K370S:S364K;T411T / E360E / Q362E:D401K;L368D / K370S:S364K / E357L;K370S:S364K / E357Q;T366S / L368A / Y407V:T366W and T366S / L368A / Y407V / Y349C:T366W / S354C (EU numbering)). In an exemplary embodiment, one of the first or second variant Fc domains contains the heterodimerized scuba riant L368D / K370S, and the other of the first or second variant Fc domains contains the heterodimerized scuba riant S364K / E357Q, with numbering following EU numbering.

[0306] In some embodiments, the variant Fc domain includes a de-evolution variant (including the one shown in Figure 5). In some embodiments, each of the first and second variant Fc domains includes a de-evolution variant E233P / L234V / L235A / G236_ / S267K, numbered according to EU numbering.

[0307] In some embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first monomer includes a pI variant (including those shown in Figure 4). In exemplary embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first monomer includes pI variants N208D / Q295E / N384D / Q418E / N421D, numbered according to EU numbering.

[0308] In some embodiments, the scFv of the 2+1 Fab2-scFv-Fc antibody provided herein includes a charged scFv linker (including the one shown in Figure 6). In some embodiments, the 2+1 Fab2-scFv-Fc antibody provided herein includes the FcRn variant M428L / N434S, which is numbered according to EU numbering.

[0309] In exemplary embodiments, the 2+1 Fab2-scFv-Fc form antibody includes the amino acid modification combinations shown in Figure 8. In such embodiments, the first variant Fc domain contains the heterodimerized scuba rianto L368D / K370S, the second variant Fc domain contains the heterodimerized scuba rianto S364K / E357Q, each of the first and second variant Fc domains contains the diminished variant E233P / L234V / L235A / G236_ / S267K, and the constant domain (CH1-hinge-CH2-CH3) of the first monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, with numbering following EU numbering. In some embodiments, the scFv of the 2+1 Fab2-scFv-Fc antibody provided herein includes a (GKPGS) 4-charged scFv linker (SEQ ID NO: 24). In some embodiments, the 2+1 Fab2-scFv-Fc antibody provided herein includes the FcRn variant M428L / N434S, which is numbered according to EU numbering.

[0310] In some embodiments, the CH1-hinge-CH2-CH3 of the first monomer contains the amino acid variant L368D / K370S / N208D / Q295E / N384D / Q418E / N421D / E233P / L234V / L235A / G236del / S267K, and the second Fc domain contains the amino acid variant S364K / E357Q / E233P / L234V / L235A / G236del / S267K, with the numbering following EU numbering.

[0311] In some embodiments, the scFv of the second monomer of the 2+1 Fab2-scFv-Fc antibody is CD28-binding, and the VH1 of the first and second monomers and the VL1 of the common light chain each form a binding domain that binds to TROP2. The 2+1 Fab2-scFv-Fc antibody of interest may include any preferred CD28-binding domain, including any of the CD28-binding domains provided herein. In some embodiments, the CD28 binding domain is one of the following CD28 binding domains or variants thereof: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1.71, CD28.3[CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28]_H 0L0, TGN1412_H1L1, 341VL34[CD28]_H1L1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9.3[CD28]_H0L0, hu9.3[CD28]_H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, and TN228[CD28]_H4L2 (Figures 16, 19, and 22).

[0312] In some embodiments of the 2+1 Fab2-scFv-Fc format, the anti-CD28 ABD has VH and VL domains selected from the following:

[0313] (i) VH containing vhCDR1, vhCDR2, and vhCDR3, wherein each VH has the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of any VH or variant thereof shown in Figures 16, 19, and 22; and (ii) VL containing vlCDR1, vlCDR2, and vlCDR3, wherein each VL has the amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of any VL or variant thereof shown in Figures 16, 19, and 22; or

[0314] (i) a VH containing vhCDR1, vhCDR2, and vhCDR3, each having the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of the VH or its variant shown in Figures 16, 17, or 34; and (ii) a VL containing vlCDR1, vlCDR2, and vlCDR3, each having the amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of the VL or its variant shown in Figures 16, 18, or 36.

[0315] In some embodiments of the 1+1 Fab-scFv-Fc format, the anti-CD28 ABD has VH and VL domains selected from the following:

[0316] (i) a VH having the amino acid sequence of any VH or a variant thereof of the CD28 binding domain shown in Figures 16, 19, and 22, and (ii) a VL having the amino acid sequence of any VL or a variant thereof of the CD28 binding domain shown in Figures 16, 19, and 22; or

[0317] (i) VH having the amino acid sequence of VH or a variant thereof shown in Figure 16, 17, or 34, and (ii) VL having the amino acid sequence of VL or a variant thereof shown in Figure 16, 18, or 36.

[0318] In some embodiments, the VH1 of the first and second monomers and the VL1 of the common light chain of the 2+1 Fab2-scFv-Fc antibody each form a TROP2-binding domain. The 2+1 Fab2-scFv-Fc antibody of interest may include any suitable TROP2-binding domain, including any of the TROP2-binding domains provided herein. Exemplary TROP2-binding domains that can be used in the 2+1 Fab2-scFv-Fc antibody of interest are shown in Figure 23. Additional VH and VL sequences of exemplary TROP2-binding domains that can be used in the 2+1 Fab2-scFv-Fc antibody of interest are shown in Figures 24-28, 36, and 52-56.

[0319] In some embodiments of the 2+1 Fab2-scFv-Fc format, the anti-TROP2 ABD each includes VH and VL domains selected from the following:

[0320] (i) a VH having the amino acid sequence of any one of the TROP2 binding domains shown in Figure 23 or a variant thereof, and (ii) a VL having the amino acid sequence of the TROP2 binding domain or a variant thereof shown in Figure 23; or

[0321] (i) VH having the amino acid sequence of VH or its variant shown in Figures 24-28, and (ii) VL having the amino acid sequence of VL or its variant VL shown in Figures 24-28.

[0322] Figure 10 shows several exemplary Fc domain sequences useful in 2+1 Fab2-scFv-Fc format antibodies. Figures 12 and 13 also provide exemplary CH1-hinge domains, CH1 domains, and hinge domains that may be included in the second or third monomer of the 2+1 Fab2-scFv-Fc format. Furthermore, Figure 14 provides useful CL sequences that may be used in this format.

[0323] 6.2+1 Fab2-Fc×scFv-Fc format One heterodimer antibody configuration particularly utilized in the target anti-CD28 × anti-TROP2 antibodies provided herein is the 2+1 Fab2-Fc × scFv-Fc configuration (also known as the “stacked bottle opener”) shown in Figure 15N. This configuration comprises a first monomer, a second monomer, and two common light chains. The first monomer contains VH1-CH1-domain linker-VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, where VH1 is the first variable weight domain and CH2-CH3 is the first variant Fc domain. The domain linker can be any useful domain linker (see, for example, Figure 7). The second monomer contains a single-stranded Fv ("scFv") covalently linked to the second variant Fc domain by the domain linker (scFv-domain linker-CH2-CH3). Each of the two common light chains contains VL1-CL, where VL1 is a first variable light domain. The scFv of the second monomer contains a second variable heavy domain (VH2) linked to the second variable light domain (VL2) by an scFv linker. In this embodiment, the two VH1-CH1 of the first monomer each interact with one of the two common light chains to form two identical first antigen-binding domains, and VH2 and VL2 form a second antigen-binding domain. In some embodiments, each of the first antigen-binding domains is a TROP2-binding domain, and the second antigen-binding domain is a CD28-binding domain. As with many of the embodiments herein, these constructs may include scuba riants, pI variants, attenuation variants, additional Fc variants, etc., as desired and as described herein.

[0324] In some embodiments, the second and second Fc domains of the 2+1 Fab2-Fc×scFv-Fc antibody are variant Fc domains containing heterodimerized scuba riants (e.g., the set of amino acid substitutions shown in Figures 3 and 9). Particularly useful heterodimerized scuba riants include S364K / E357Q:L368D / K370S;L368D / K370S:S364K;L368E / K370S:S364K;T411T / E360E / Q362E:D401K;L368D / K370S:S364K / E357L;K370S:S364K / E357Q;T366S / L368A / Y407V:T366W and T366S / L368A / Y407V / Y349C:T366W / S354C (EU numbering)). In an exemplary embodiment, one of the first or second variant Fc domains contains the heterodimerized scuba riant L368D / K370S, and the other of the first or second variant Fc domains contains the heterodimerized scuba riant S364K / E357Q, with numbering following EU numbering.

[0325] In some embodiments, the variant Fc domain includes a de-evolution variant (including the one shown in Figure 5). In some embodiments, each of the first and second variant Fc domains includes a de-evolution variant E233P / L234V / L235A / G236_ / S267K, numbered according to EU numbering.

[0326] In some embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first monomer includes a pI variant (including those shown in Figure 4). In exemplary embodiments, the steady-state domain (CH1-hinge-CH2-CH3) of the first monomer includes pI variants N208D / Q295E / N384D / Q418E / N421D, numbered according to EU numbering.

[0327] In some embodiments, the 2+1 Fab2-Fc×scFv-Fc antibody provided herein includes a charged scFv linker (including the one shown in Figure 6). In some embodiments, the 2+1 Fab2-scFv-Fc antibody provided herein includes the FcRn variant M428L / N434S, which is numbered according to EU numbering.

[0328] In exemplary embodiments, the 2+1 Fab2-Fc×scFv-Fc antibody includes the amino acid modification combinations shown in Figure 8. In such embodiments, the first variant Fc domain contains the heterodimerized scuba rianto L368D / K370S, the second variant Fc domain contains the heterodimerized scuba rianto S364K / E357Q, each of the first and second variant Fc domains contains the diminished variant E233P / L234V / L235A / G236_ / S267K, and the constant domain (CH1-hinge-CH2-CH3) of the first monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, with numbering following EU numbering. In some embodiments, the scFv of the 2+1 Fab2-scFv-Fc antibody provided herein includes a (GKPGS) 4-charged scFv linker (SEQ ID NO: 24). In some embodiments, the 2+1 Fab2-scFv-Fc antibody provided herein includes the FcRn variant M428L / N434S, which is numbered according to EU numbering.

[0329] In some embodiments, the CH1-hinge-CH2-CH3 of the first monomer contains the amino acid variant L368D / K370S / N208D / Q295E / N384D / Q418E / N421D / E233P / L234V / L235A / G236del / S267K, and the second Fc domain contains the amino acid variant S364K / E357Q / E233P / L234V / L235A / G236del / S267K, with the numbering following EU numbering.

[0330] In some embodiments, the scFv of the second monomer of the 2+1 Fab2-Fc×scFv-Fc antibody is CD28-binding, and the VH1 of the first and second monomers and the VL1 of the common light chain each form a binding domain that binds to TROP2. The 2+1 Fab2-scFv-Fc antibody of interest may contain any preferred CD28-binding domain, including any of the CD28-binding domains provided herein. In some embodiments, the CD28 binding domain is one of the following CD28 binding domains or variants thereof: 1A7[CD28]_H1L1, 1A7[CD28]_H1.1_L1, 1A7[CD28]_H1_L1.71, 1A7[CD28]_H1.1_L1.71, 1A7[CD28]_H1.14_L1, 1A7[CD28]_H1.14_L1.71, CD28.3[CD28]_H0L0, hCD28.3[CD28]_H1L1, 5.11A1[CD28]_H 0L0, TGN1412_H1L1, 341VL34[CD28]_H1L1, 341VL36[CD28]_H1L1, 281VL4[CD28]_H1L1, HuTN228[CD28]_H1L1, PV1[CD28]_H0L0, m9.3[CD28]_H0L0, hu9.3[CD28]_H1L1, 9G2[CD28]_H0L0, 9G2[CD28]_H1L1, 2F10A3.140[CD28]_H1L1, and TN228[CD28]_H4L2 (Figures 16, 19, and 22).

[0331] In some embodiments of the 2+1 Fab2-Fc×scFv-Fc format, the anti-CD28 ABD has VH and VL domains selected from the following:

[0332] (i) VH containing vhCDR1, vhCDR2, and vhCDR3, wherein each VH has the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of any VH or variant thereof shown in Figures 16, 19, and 22; and (ii) VL containing vlCDR1, vlCDR2, and vlCDR3, wherein each VL has the amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of any VL or variant thereof shown in Figures 16, 19, and 22; or

[0333] (i) a VH containing vhCDR1, vhCDR2, and vhCDR3, each having the amino acid sequences of vhCDR1, vhCDR2, and vhCDR3 of the VH or its variant shown in Figures 16, 17, or 34; and (ii) a VL containing vlCDR1, vlCDR2, and vlCDR3, each having the amino acid sequences of vlCDR1, vlCDR2, and vlCDR3 of the VL or its variant shown in Figures 16, 18, or 36.

[0334] In some embodiments of 2+1 Fab2-Fc×scFv-Fc, the anti-CD28 ABD has VH and VL domains selected from the following:

[0335] (i) a VH having the amino acid sequence of any VH or a variant thereof of the CD28 binding domain shown in Figures 16, 19, and 22, and (ii) a VL having the amino acid sequence of any VL or a variant thereof of the CD28 binding domain shown in Figures 16, 19, and 22; or

[0336] (i) VH having the amino acid sequence of VH or a variant thereof shown in Figure 16, 17, or 34, and (ii) VL having the amino acid sequence of VL or a variant thereof shown in Figure 16, 18, or 36.

[0337] In some embodiments, the VH1 of the first and second monomers and the VL1 of the common light chain of the 2+1 Fab2-Fc×scFv-Fc antibody each form a TROP2-binding domain. The 2+1 Fab2-Fc×scFv-Fc antibody of interest may include any suitable TROP2-binding domain, including any of the TROP2-binding domains provided herein. Exemplary TROP2-binding domains that can be used in the 2+1 Fab2-Fc×scFv-Fc antibody of interest are shown in Figure 23. Additional VH and VL sequences of exemplary TROP2-binding domains that can be used in the 2+1 Fab2-Fc×scFv-Fc antibody of interest are shown in Figures 24-28 and 52-56.

[0338] In some embodiments of the 2+1 Fab2-Fc×scFv-Fc format, the anti-TROP2 ABD each includes VH and VL domains selected from the following:

[0339] (i) a VH having the amino acid sequence of any one of the TROP2 binding domains shown in Figure 23 or a variant thereof, and (ii) a VL having the amino acid sequence of the TROP2 binding domain or a variant thereof shown in Figure 23; or

[0340] (i) VH having the amino acid sequence of VH or its variant shown in Figures 24-28, and (ii) VL having the amino acid sequence of VL or its variant VL shown in Figures 24-28.

[0341] Figure 10 shows several exemplary Fc domain sequences useful in 2+1 Fab2-Fc×scFv-Fc format antibodies. Figures 12 and 13 provide exemplary CH1-hinge domains, CH1 domains, and hinge domains that may be included in the second or third monomer of the 2+1 Fab2-Fc×scFv-Fc format. Furthermore, Figure 14 provides useful CL sequences that may be used in this format.

[0342] 7. Dual scFv format One heterodimer antibody form particularly used in the target bispecific anti-CD28 × anti-TROP2 antibody is known in the art and is the bispecific scFv form shown in Figure 15F. In this embodiment, the heterodimer bispecific antibody consists of two scFv-Fc monomers (both in either the (vh-scFv linker-vl-[any domain linker]-CH2-CH3) form or the (vl-scFv linker-vh-[any domain linker]-CH2-CH3) form, or one monomer in one orientation and the other in the other orientation).

[0343] In this case, all ABDs are in scFv format. A dual scFv anti-CD28 × anti-TROP2 antibody may contain any suitable TROP2-binding domain and CD28-binding domain, including either the TROP2-binding domain or the CD28-binding domain provided herein.

[0344] In addition, the Fc domain in the double scFv form is a scuba rian (for example, a set of amino acid substitutions shown in Figures 3 and 9, with particularly useful scuba rians being S364K / E357Q:L368D / K370S;L368D / K370S:S364K;L368E / K370S:S364K;T411T / E360E / Q362E:D401K;L368D / K370S:S364K / E357 The linker is selected from the group consisting of L, K370S:S364K / E357Q, T366S / L368A / Y407V:T366W, and T366S / L368A / Y407V / Y349C:T366W / S354C), optionally includes a de-escalating variant (including those shown in Figure 5), optionally includes a charged scFv linker (including those shown in Figure 6), and the heavy chain includes a pI variant (including those shown in Figure 4).

[0345] In some embodiments, the dual scFv form includes a scuba riant, a pI variant, and a decay variant. Thus, in some embodiments, a) a first monomer (VH1-scFv linker-VL1-[any domain linker]-CH2-CH3 or VL1-scFv linker-VH1-[any domain linker]-CH2-C) containing a scuba riant S364K / E357Q, a decay variant E233P / L234V / L235A / G236del / S267K, and an scFv that binds the first antigen. H3) and b) include forms comprising a first monomer (VH1-scFv linker-VL1-[any domain linker]-CH2-CH3 or VL1-scFv linker-VH1-[any domain linker]-CH2-CH3) containing the scuba antigen L368D / K370S, the diminished variant E233P / L234V / L235A / G236del / S267K, and scFv for binding a second antigen. pI variants may be as outlined herein, but the most common are charged scFv linkers with opposite charges for each monomer. FcRn variants, in particular 428L / 434S, may optionally be included.

[0346] The dual scFv format may include any preferred TROP2-binding domain and CD28-binding domain, including the TROP2-binding domain and CD28-binding domain provided herein, as well as any of the associated VH and VL or their variants (see, for example, Figures 16-19, 22-28, 34, 36 and 52-56).

[0347] 8.1 Arm Center - scFv One heterodimer antibody form particularly utilized in the anti-CD28 × anti-TROP2 antibodies provided herein is the one-arm central-scFv form shown in Figure 15J. In this embodiment, one monomer simply contains an Fc domain, while the other monomer contains a Fab domain (first antigen-binding domain), an scFv domain (second antigen-binding domain), and an Fc domain, with the scFv domain inserted between the two Fc domains.

[0348] In this embodiment, one monomer comprises a first heavy chain containing a first variable heavy domain, a CH1 domain, and an Fc domain, and scFv comprises an scFv variable light domain, an scFv linker, and an scFv variable heavy domain. scFv is covalently linked between the C-terminus of the CH1 domain of the heavy steady domain and the N-terminus of the first Fc domain using a domain linker in either orientation, i.e., VH1-CH1-[any domain linker]-VH2-scFv linker-VL2-[any domain linker]-CH2-CH3 or VH1-CH1-[any domain linker]-VL2-scFv linker-VH2-[any domain linker]-CH2-CH3. The second monomer comprises an Fc domain (CH2-CH3). This embodiment further utilizes a light chain containing a variable light domain and a steady light domain that cooperate with the heavy chain to form a Fab. As with many of the embodiments herein, these constructs may include, as desired and as described herein, scuba riants, pI variants, attenuation variants, additional Fc variants, etc.

[0349] The 1-arm central-scFv format may include any preferred TROP2-binding domain and CD28-binding domain, including the TROP2-binding domain and CD28-binding domain provided herein, as well as any of the associated VH and VL or their variants (see, for example, Figures 16-19, 22-28, 34, 36, and 52-56).

[0350] 9.1 ARM scFv-mAb format One heterodimer antibody form particularly utilized in the anti-CD28 × anti-TROP2 antibodies provided herein is the one-arm mAb-scFv form shown in Figure 15G. This form comprises 1) a first monomer containing an "empty" Fc domain, 2) a second monomer containing a first variable weight domain (VH), an scFv domain (second antigen-binding domain), and an Fc domain (the scFv domain is bound to the N-terminus of the first variable weight domain), and 3) a light chain containing a first variable light domain and a constant light domain. The first variable weight domain and the first variable light domain form the first antigen-binding domain, and scFv is the second antigen-binding domain. In this form, one of the first and second antigen-binding domains binds to CD28, and the other antigen-binding domain binds to TROP2. As with many of the embodiments herein, these constructs may include, as desired and as described herein, scuba riants, pI variants, attenuation variants, additional Fc variants, etc.

[0351] A one-arm scFv-mAb antibody may contain any preferred TROP2-binding domain and CD28-binding domain, including the TROP2-binding domain and CD28-binding domain provided herein, as well as any of the associated VH and VL variants or their variants (see, for example, Figures 16-19, 22-28, 34, 36, and 52-56).

[0352] 10.scFv-mAb format One heterodimer antibody form particularly utilized in the anti-CD28 × anti-TROP2 antibodies provided herein is the mAb-scFv form shown in Figure 15H. In this embodiment, the form relies on the use of N-terminal conjugation of scFv to one of the monomers, thereby forming a third antigen-binding domain, where the Fab portions of the two monomers each bind to one target and the "added" scFv domains bind to different targets.

[0353] In this embodiment, the first monomer comprises a first heavy chain (including a variable heavy domain and a steady domain), and the scFv variable light domain, scFv linker and scFv variable heavy domain are in either orientation ((vh1-scFv linker-vl1-[any domain linker]-vh2-CH1-hinge-CH2-CH3) or (for scFv with the opposite orientation) ((vl1-scFv linker-vh1-[any domain linker]-vh2-CH1-hinge-CH2-CH3)) It has an N-terminal covalent bond scFv. The second monomer contains a heavy chain VH2-CH1-hinge-CH2-CH3. This embodiment further utilizes a common light chain containing a variable light domain and a constant light domain that cooperate with the heavy chain to form two identical Fabs. As with many of the embodiments herein, these constructs include scuba variants, pI variants, decay variants, additional Fc variants, etc., as desired and as described herein.

[0354] The scFv-mAb antibody format may include any preferred TROP2-binding domain and CD28-binding domain, including the TROP2-binding domain and CD28-binding domain provided herein, as well as any of the associated VH and VL variants or their variants (see, for example, Figures 16-19, 22-28, 34, 36, and 52-56).

[0355] 11.mAb-Fv format One heterodimer antibody form particularly utilized in the anti-CD28 × anti-TROP2 antibodies provided herein is the mAb-Fv form (Figure 15K). In this embodiment, the form relies on the use of C-terminal attachment of an “additional” variable heavy domain to one monomer and C-terminal attachment of an “additional” variable light domain to the other monomer, thereby forming a third antigen-binding domain (i.e., an “additional” Fv domain), where the Fab portions of the two monomers bind to CD28 and the “additional” Fv domain binds to TROP2.

[0356] In this embodiment, the first monomer comprises a first heavy chain comprising a first variable heavy domain and a first steady heavy domain containing a first Fc domain, and having a first variable light domain covalently bonded to the C-terminus of the first Fc domain using a domain linker (vh1-CH1-hinge-CH2-CH3-[any linker]-vl2). The second monomer comprises a second variable heavy domain, a second steady heavy domain containing a second Fc domain, and a third variable heavy domain covalently bonded to the C-terminus of the second Fc domain using a domain linker (vh1-CH1-hinge-CH2-CH3-[any linker]-vh2). This embodiment further utilizes a common light chain comprising a variable light domain and a steady light domain, which associates with the heavy chain to form two identical Fabs containing two identical Fvs. The two C-terminally bonded variable domains constitute an “additional” third Fv. As with many of the embodiments herein, these constructs may include, as desired and as described herein, scuba riants, pI variants, attenuation variants, additional Fc variants, etc.

[0357] The mAb-Fv format may include any preferred TROP2-binding domain and CD28-binding domain, including any of the TROP2-binding domains and CD28-binding domains provided herein, as well as any of the associated VH and VL or their variants (see, for example, Figures 16-19, 22-28, 34, 36, and 52-56).

[0358] 12.Central-Fv format One heterodimer antibody form particularly utilized in the anti-CD28 × anti-TROP2 antibody provided herein is the central-scFv form shown in Figure 15L. In this embodiment, the form relies on the use of an inserted Fv domain, thereby forming an “additional” third antigen-binding domain, where the Fab portions of the two monomers bind to TROP2 and the “additional” central-Fv domain binds to CD28. The Fv domain is inserted between the Fc domain and the CH1-Fv region of the monomer, thereby providing a third antigen-binding domain, and each monomer contains components of Fv (for example, one monomer contains a variable weight domain of the “additional” central-Fv domain and the other contains a variable light domain).

[0359] In this embodiment, one monomer comprises a first heavy chain containing a first variable heavy domain, a CH1 domain, and an Fc domain and an additional variable light domain. The additional variable light domain is covalently linked between the C-terminus of the CH1 domain of the heavy steady domain and the N-terminus of the first Fc domain using a domain linker (vh1-CH1-[any linker]-vh2-hinge-CH2-CH3). The other monomer comprises a first heavy chain containing a first variable heavy domain, a CH1 domain, and an Fc domain and an additional variable heavy domain (vh1-CH1-[any linker]-vh2-hinge-CH2-CH3). The additional variable heavy domain is covalently linked between the C-terminus of the CH1 domain of the heavy chain steady domain and the N-terminus of the first Fc domain using a domain linker. This embodiment utilizes a common light chain containing a variable light chain domain and a steady light chain domain, which associates with the heavy chain to form two identical Fabs, each of which binds to TROP2. Additional variable heavy domains and additional variable light domains form an “additional” central Fv that binds to CD28. As with many of the embodiments herein, these constructs include, optionally and as described herein, scuba variants, pI variants, attenuation variants, additional Fc variants, etc.

[0360] The central-Fv format may include any preferred TROP2-binding domain and CD28-binding domain, including any of the TROP2-binding domains and CD28-binding domains provided herein, as well as any of the associated VH and VL or their variants (see, for example, Figures 16-19, 22-28, 34, 36, and 52-56).

[0361] 13. Non-heterodimer bispecific antibodies As will be understood by those skilled in the art, the anti-CD28 × anti-TROP2 antibodies provided herein may also include a non-heterodimerative bispecific form (Figure 15I). In this form, the anti-CD28 × anti-TROP2 comprises: 1) a first monomer containing VH1-CH1-hinge-CH2-CH3; 2) a second monomer containing VH2-CH1-hinge-CH2-CH3; 3) a first light chain containing VL1-CL; and 4) a second light chain containing VL2-CL. In such embodiments, VH1 and VL1 form a first antigen-binding domain, and VH2 and VL2 form a second antigen-binding domain. One of the first or second antigen-binding domains binds to CD28, and the other antigen-binding domain binds to TROP2.

[0362] Anti-CD28 × anti-TROP2 antibodies in the form of non-heterodimerary bispecific antibodies may include any preferred TROP2-binding domain and CD28-binding domain, including the TROP2-binding domain and CD28-binding domain provided herein, as well as any of the associated VH and VL or their variants (see, for example, Figures 16-19, 22-28, 34, 36, and 52-56).

[0363] 14. Trident format In some embodiments, the anti-CD28 × anti-TROP2 antibodies provided herein are in the “Trident” form, as commonly described in WO2015 / 184203 (the whole and in particular the figures, legend, definitions, and sequences of the “heterodimerization-promoting domain” or “HPD” containing the “K-coil” and “E-coil” sequences are incorporated herein by express reference). The Trident relies on the use of two different HPDs that associate to form a heterodimer structure as structural components. See Figure 15M. In this embodiment, the trident form comprises a “conventional” heavy and light chain (e.g., VH1-CH1-hinge-CH2-CH3 and VL1-CL), a third chain (VH2-(linker)-VL3-HPD1) containing a first “diabody-type binding domain” or “DART®”, and a fourth chain (VH3-(linker)-(linker)-VL2-HPD2) containing a second DART®. VH1 and VL1 form the first ABD, VH2 and VL2 form the second ABD, and VH3 and VL3 form the third ABD. In some cases, as shown in Figure 15M, the second and third ABDs bind to the same antigen.

[0364] The Trident format may include any preferred TROP2-binding domain and CD28-binding domain, including the TROP2-binding domain and CD28-binding domain provided herein, as well as any of the associated VH and VL or their variants (see, for example, Figures 16-19, 22-28, 34, 36, and 52-56).

[0365] V. Nucleic acids In another embodiment, what is provided herein is a nucleic acid composition encoding an anti-CD28 × anti-TROP2 antibody provided herein. The nucleic acid composition may refer to one or more polynucleotides.

[0366] As will be understood by those skilled in the art, nucleic acid compositions will depend on the form and backbone of the heterodimeric protein. Therefore, for example, if the form requires three amino acid sequences for 1+1 CLC, 2+1 CLC, 1+1 Fab-scFv-Fc, 2+1 mAb-scFv, or 2+1 Fab2-scFv-Fc forms, then three polynucleotides may be incorporated into one or more expression vectors for expression. In exemplary embodiments, each polynucleotide is incorporated into a different expression vector.

[0367] As is known in the art, the nucleic acids encoding the binding domains and antibody components disclosed herein may be incorporated into an expression vector as is known in the art, and depending on the host cell used to form the heterodimeric antibody of the present invention. Typically, the nucleic acids are functionally ligated to any number of regulatory elements (promoters, origins of replication, selection markers, ribosome binding sites, inducers, etc.). The expression vector may be extrachromosomal or embedded.

[0368] The polynucleotide and / or expression vector of the present invention is then transformed into any number of different types of host cells well known in the art, including mammalian, bacterial, yeast, insect, and / or fungal cells (mammalian cells (e.g., CHO cells) are used in many embodiments).

[0369] In some embodiments, each polynucleotide encoding a monomer is contained within a single expression vector, typically under different or the same promoter control. In certain embodiments of use in the present invention, each of these polynucleotides is contained in a different expression vector. Different vector ratios can be used to induce heterodimerization, as shown herein and in US62 / 025,931 (incorporated herein by reference). Surprisingly, the protein contains the first monomer:second monomer:light chain (in many embodiments herein having three polypeptides including a heterodimeric antibody) in a 1:1:2 ratio, while these are not the ratios that yield the best results.

[0370] The antibodies provided herein are produced by culturing host cells containing an expression vector(s), as is well known in the art. Once produced, conventional antibody purification steps, including an ion-exchange chromatography step, are performed. As discussed herein, by making the pIs of the two monomers differ by at least 0.5, separation by ion-exchange chromatography, isoelectric focusing, or other isoelectric-sensitive methods may be possible. That is, the inclusion of pI substitution modifies the isoelectric point (pI) of each monomer so that each monomer has a different pI and the heterodimer also has a distinguishable pI, thus facilitating isoelectric purification (e.g., anion-exchange column, cation-exchange column) of the "1+1 Fab-scFv-Fc" heterodimer. These substitutions are also useful in determining and monitoring any contaminated double scFv-Fc and mAb homodimers after purification (e.g., IEF gel, cIEF, and analytical IEX column).

[0371] VI. Biological and biochemical functions of anti-CD28 × anti-TROP2 antibodies Typically, the anti-CD28 × anti-TROP2 antibodies described herein are administered to patients with TROP2-associated cancers, and their efficacy is evaluated using the numerous methods described herein. Therefore, standard assays for efficacy, such as assessment of cancer burden, tumor size, and the presence or extent of metastasis, may be performed, and cancer immunotherapy may also be evaluated based on an assessment of the immune status. This can be done using numerous methods, including both in vitro and in vivo assays.

[0372] A. Antibody composition for in vivo administration Embodiments of the present invention relate to pharmaceutical compositions comprising one of the anti-CD28 × anti-TROP2 antibodies described herein and a pharmaceutically acceptable carrier. The anti-CD28 × anti-TROP2 antibody formulations described herein are prepared for storage in the form of lyophilized formulations or aqueous solutions by mixing the antibody of desired purity with any pharmaceutically acceptable carrier, excipient, or stabilizer (Remington's Pharmaceutical Sciences 16th edition, Osol, A. Ed.

[1980] ). Acceptable carriers, excipients, or stabilizers are non-toxic to the recipient at the dosage and concentration used and include buffers, e.g., phosphoric acid, citrate, and other organic acids; antioxidants, including ascorbic acid and methionine; preservatives (e.g., octadecyldimethylbenzylammonium chloride; hexamethonium chloride; benzalkonium chloride, benzethonium chloride; phenol, butyl alcohol, or benzyl alcohol; alkylparabens, e.g., methylparaben or propylparaben; catechol; resorcinol; cyclohexanol; 3-pentanol; and m-cresol); low molecular weight (less than about 10 residues) polypeptides; proteins, e.g. These include serum albumin, gelatin, or immunoglobulin; hydrophilic polymers, e.g., polyvinylpyrrolidone; amino acids, e.g., glycine, glutamine, asparagine, histidine, arginine, or lysine; monosaccharides, disaccharides, and other carbohydrates, including glucose, mannose, or dextrin; chelating agents, e.g., EDTA; sugars, e.g., sucrose, mannitol, trehalose, or sorbitol; salt-forming counterions, e.g., sodium; metal complexes (e.g., Zn-protein complexes); and / or nonionic surfactants, e.g., TWEEN®, PLURONICS®, or polyethylene glycol (PEG).

[0373] VII. Treatment The compositions of the present invention, once prepared, are utilized in numerous oncological applications, particularly when used in conjunction with anticancer therapies such as antitumor bispecific antibodies, by treating cancer, typically by enhancing the immune response (e.g., T cell activation and proliferation). In some embodiments, the antibodies provided herein enhance the immune response (e.g., T cell activation and proliferation) by providing agonist-like costimulation to T cells in the tumor microenvironment expressing TROP2.

[0374] In some embodiments, the anti-CD28 × anti-TROP2 bispecific antibody provided herein is administered in conjunction with an antitumor therapy, for example, an antitumor-associated antigen (TAA) bispecific antibody.

[0375] A. Anti-CD28 x anti-TROP2 / anti-TROP2 bispecific antibody In some embodiments, the anti-CD28 × anti-TROP2 antibody provided herein is administered together with an anti-TROP2 bispecific antibody, which is a T cell engagement bispecific antibody, such as one that binds to human CD3.

[0376] In classical T cell / APC interactions, there are two signals: a first signal (signal 1) provided by TCR reactivity with peptide-MHC, and a second signal (signal 2) provided by crosslinking of CD28 by CD80 / CD86 expressed on APCs. These together fully activate T cells (see Figure 29A). In contrast, treatment with a CD3 bispecific antibody targeting tumor-associated antigens (TAAs) (i.e., an anti-CD3 × anti-TROP2 bispecific antibody) provides only the first signal.

[0377] Without being bound by any particular theory of operation, the anti-CD28 × anti-TROP2 bispecific antibody provided herein is thought to be able to enhance the antitumor response of the anti-CD3 × anti-TROP2 bispecific antibody through CD28 co-stimulation (see Figure 29B). Thus, in one embodiment, what is provided herein is a method for treating TROP2-related cancer in a patient by administering to the patient an anti-CD3 × anti-TROP2 bispecific antibody and the anti-CD28 × anti-TROP2 bispecific antibody provided herein.

[0378] Anti-CD3 × anti-TROP2 antibodies useful for producing “Signal 1” when combined with the target anti-CD28 × anti-TROP2 antibody include those having either a CD3-binding domain or a TROP2-binding domain (see, for example, Figures 23-28 and 32) as provided herein. Suitable antibody forms for such anti-CD3 × anti-TROP2 antibodies include, but are not limited to, the antibody forms described herein (see, for example, Figure 15). In some embodiments, the anti-CD3 × anti-TROP2 antibody and the anti-CD28 × anti-TROP2 antibody used in combination bind to the same TROP2 epitope. In some embodiments, the anti-CD3 × anti-TROP2 antibody and the anti-CD28 × anti-TROP2 antibody used in combination bind to different TROP2 epitopes.

[0379] B. Dosage Modalities The antibodies provided herein are administered to subjects by known methods, for example, intravenous administration as a bolus or by continuous infusion over a predetermined period.

[0380] C. Treatment Modalities In the method of the present invention, the therapy is used to provide a positive therapeutic response with respect to a disease or condition.

[0381] A “positive therapeutic response” is intended to be an improvement in the disease or condition, and / or an improvement in symptoms associated with the disease or condition. For example, a positive therapeutic response would refer to one or more of the following improvements in the disease: (1) a reduction in the number of neoplastic cells; (2) an increase in neoplastic cell death; (3) an inhibition of neoplastic cell survival; (5) an inhibition of tumor growth (i.e., slowing it down to some extent, preferably stopping it); (6) an increased patient survival rate; and (7) some relief of one or more symptoms associated with the disease or condition.

[0382] A positive therapeutic response in any given disease or condition can be determined by standardized response criteria specific to that disease or condition. Tumor response can be assessed for changes in tumor morphology (i.e., overall tumor load, tumor size, etc.) using screening techniques such as magnetic resonance imaging (MRI) scans, X-ray imaging, computed tomography (CT) scans, bone scan imaging, endoscopy, and tumor biopsy sampling including bone marrow aspiration (BMA) and counting of circulating tumor cells.

[0383] In addition to these positive therapeutic responses, those receiving therapy may experience beneficial effects such as improvement in disease-related symptoms.

[0384] The treatment according to the present invention includes the “therapeutic effective dose” of the drug used. The “therapeutic effective dose” refers to the amount that is effective in the dosage and duration required to achieve the desired therapeutic outcome.

[0385] The therapeutically effective dose may vary depending on factors such as the individual's disease state, age, sex, and weight, as well as the drug's ability to induce the desired response in the individual. The therapeutically effective dose is also defined as the amount at which the therapeutically beneficial effect outweighs any toxic or adverse effects of the antibody or antibody portion.

[0386] The "therapeutic effective dose" for tumor therapy can also be measured by its ability to stabilize disease progression. The ability of a compound to inhibit cancer can be evaluated in animal model systems to predict efficacy in human tumors.

[0387] Alternatively, this property of a composition can be assessed by testing the compound's ability to inhibit cell growth or induce apoptosis using in vitro assays known to those skilled in the art. A therapeutically effective amount of a therapeutic compound may reduce tumor size or otherwise improve symptoms in the subject. Those skilled in the art may determine such an amount based on factors such as the size of the subject, the severity of the subject's symptoms, and the specific composition or route of administration selected.

[0388] The drug regimen may be adjusted to provide the optimal desired response (e.g., therapeutic response). For example, a single bolus may be administered, several divided doses may be administered over time, or the dose may be proportionally reduced or increased as indicated by the requirements of the therapeutic situation. Parenteral compositions may be formulated in drug unit form for ease of administration and uniformity of dosage. When used herein, drug unit form refers to a physically distinct unit suitable as a unit dose for the subject being treated; each unit contains a predetermined amount of the active compound calculated to produce the desired therapeutic effect in relation to the required pharmaceutical carrier.

[0389] The specifications for the drug unit form of the present invention are determined by and directly depend on (a) the unique characteristics of the active compound and the specific therapeutic effect to be achieved, and (b) the limitations inherent in the art to formulate such an active compound for highly sensitive treatment in an individual.

[0390] The efficient dosage and administration regimen for the bispecific antibody used in the present invention depend on the disease or condition being treated and can be determined by those skilled in the art. [Examples]

[0391] Example 1: CD28-binding domain The sequences of human, mouse, and cynomolgus monkey CD28 are shown in Figures 1A and 1B, which are useful for developing cross-reactive CD28 antigen-binding domains to facilitate clinical development.

[0392] 1A: Novel CD28-bound domain An approach considered to avoid the superagonism associated with TGN1412 was to generate novel CD28-binding domains with lower affinity binding to CD28 than TGN1412 and / or binding to different CD28 epitopes than TGN1412. One means of generating such novel CD28-binding domains was to pan a new in-house phage library against CD28.

[0393] 1A(a): Phage-derived clone 1A7 It should be noted that this phage library utilizes diverse human germline VL introduced into LCDR3. The amino acid sequence of an exemplary phage-derived clone 1A7 is shown in Figure 16.

[0394] To optimize CD28 bsAb, numerous 1A7 affinity variants were developed by manipulating VH variants (exemplary sequences shown in Figures 17A-17F and SEQ ID NOs. 117-185), VL variants (exemplary sequences shown in Figures 18A-18I and SEQ ID NOs. 186-291), the consensus sequence shown in Figures 20A-20B, and combinations thereof (exemplary sequences shown in Figures 19A-19C). The monovalent affinity of the exemplary variants in relation to scFv is shown in Figure 21.

[0395] Furthermore, there was a request to utilize the 1A7 binding domain in a common light chain format. The phage library for discovering CD28 clone 1A7 used the same parental human germline VL as the phage library for discovering the TROP2 binding domain 1F11-1A3.315 described in Example 2A. Surprisingly, the variable light domain of clone 1A7 was found to differ from the parental germline VL by only one amino acid in LCDR3. However, despite this single amino acid difference, the VH of 1A7 did not adequately pair with the parental germline VL, resulting in reduced binding. Therefore, an affinity-optimized library was generated focusing only on substitutions in the variable heavy domain of 1A7 (to ensure that the germline VL's ability to function as a common light chain with the TROP2 binding domain of Example 2A was not compromised). The amino acid sequences of exemplary affinity-optimized 1A7 variable domains are shown in Figure 34 and sequence numbers XXX-YYY. The monovalent KD of the affinity-manipulated 1A7 Fab variant was determined by Octet, and the data is shown in Figure 35. It should be noted that the values ​​may vary between experiments and batches. For example, in another series of experiments, the 1A7_H1.129_IGKV1-39_L1 domain showed an affinity of 60-70 nM, while the 1A7_H1.118_IGKV1-39_L1 domain showed an affinity of approximately 400 nM. However, the trend remained the same; for example, 1A7_H1.129_IGKV1-39_L1 has a stronger affinity than 1A7_H1.118_IGKV1-39_L1. Nevertheless, to illustrate the binding domains described herein, we use the KD in Figure 35.

[0396] 1B: Additional CD28 binding domains The sequences of additional CD28-binding domains that can be used in the TROP2×CD28 bsAb of the present invention are shown in Figures 22A to 22H.

[0397] Example 2: TROP2 binding domain The sequences of human, mouse, and cynomolgus monkey TROP2 are shown in Figure 2. These are useful for developing cross-reactive TROP2 antigen-binding domains that facilitate clinical development.

[0398] 2A: Novel TROP2 binding domain Since bispecific antibody formats utilizing only the Fab domain may have lower immunogenicity and stability than those utilizing the scFv domain, a common light chain format was explored to manipulate TROP2×CD28 bsAb. A phage library utilizing the diversity of stationary human germline VL (the same human germline VL as in Example 1A(a), but without any diversity) and VH was used. The amino acid sequence of an exemplary phage-derived TROP2 clone 1F11-1A3.315 is shown in Figure 36.

[0399] To adjust the binding affinity (and the efficacy of the bispecific molecule) to TROP2, the affinity ladder was manipulated (exemplary sequences are shown in Figure 36). First, the KD of the affinity-manipulated 1F11-1A3.315 variants was determined in the divalent form, and the data is shown in Figure 37. Subsequently, the monovalent KD was determined. 1F11-1A3.315_H1L1 was 13 nM, 1F11-1A3.315_H1.10_L1 was 5 nM, and 1F11-1A3.315_H1.19_L1 was 2 nM. These are weak (because these variants are monovalent), but the trend is consistent with the divalent form (i.e., H1L1

[0400] While the affinity maturation campaign described above attempted to improve the binding affinity of the 1F11-1A3.315 clone, another approach explored was weaker TROP2 binding. Therefore, additional phage-derived TROP2 clones were identified. The sequences are shown in Figures 52–56. The KDs were determined as follows: 1C9A4.313 (40 nM), 1B11A3.316 (54 nM), 1B2A4.312 (64 nM), 1A3A4.312 (80 nM), and 2C5A3.316 (345 nM).

[0401] 2B: Additional TROP2 binding domains The sequences of the TROP2 binding domains that can be used in the TROP2×CD28 bsAb of the present invention are shown in Figures 23-28. ​

[0402] Example 3: Operation of TROP2×CD28 bsAb T cells require multiple signals for complete activation and differentiation. Signal 1, facilitated by the recognition of peptide-MHC (pMHC) complexes by the T cell receptor (TCR), is absolutely essential for T cell activation. Signal 2, which acts synergistically with and amplifies signal 1, is typically provided by the interaction of CD28 itself with CD80 and CD86, which are ligands for CD28. CD28 engagement alone is typically inactive, but when combined with signal 1 activation, it promotes additional activation, survival, and proliferation signals (including IL-2 secretion) (see Figure 29). Since CD80 and CD86 are spontaneously expressed only by professional antigen-presenting cells (APCs), the degree of CD28 costimulation in the tumor environment can vary considerably. By creating this novel class of tumor-targeted CD28 bispecific antibodies, it is possible to mimic CD28's CD80 / CD86 engagement and provide an artificial source of signal 2. Notably, the signal may be provided by the tumor cell's innate TCR:pMHC recognition, or by a combination of CD28 and CD3 bispecificity (which can mimic signal 1). With these concepts in mind, we considered TROP2×CD28 bsAb and explored many forms of use. A schematic diagram is outlined in Figure 15.

[0403] 3A:1+1 Fab-scFv-Fc format One exemplary form utilizing the Fab domain and scFv is the 1+1 Fab-scFv-Fc form (schematically shown in Figure 15A), which includes a first monomer containing a single-stranded Fv ("scFv") with first antigen-binding specificity covalently bound to a first heterodimer Fc domain, i.e., scFv-domain linker-CH2-CH3; a second monomer containing a heavy chain, i.e., VH-CH1-hinge-CH2-CH3 (where CH2-CH3 is a second heterodimer Fc domain complementary to the first heterodimer Fc domain); and a light chain (LC) separately transfected so that the Fab domain with second antigen-binding specificity is formed together with a variable heavy domain. This (and other) bispecific forms can utilize any number of heterodimerization approaches known in the art, in combination with any number of approaches for purifying heterodimers (including those shown in Figure 3) from contaminated homodimers. To link the VH and VL domains of scFv, any number of linkers known in the art can be used. Finally, it may be useful to maximize the serum half-life of bsAb, and any number of half-life extension variants known in the art can also be used for these bsAb.

[0404] In particular, 1+1 Fab-scFv-Fc bsAb can utilize skeleton 1 or 11 in Figure 10. The skeleton utilizes the L368D / K370S (on HC):S364K / E357Q (on scFv-Fc) heterodimer Fc variant. The HC side further includes the pI variant N208D / Q295E / N384D / Q418E / N421D to increase the negative charge of the heavy chain. The scFv utilizes a positively charged (GKPGS)4 linker (SEQ ID NO: 24) between the VH and VL domains to increase the positive charge of the scFv-Fc chain. In summary, these two approaches enable the easy purification of heterodimers from contaminated homodimers. The FcγR reduction variants used in this platform are E233P / L234V / L235A / G236_ / S267K substitutions in both HC and scFv-Fc monomers. In some cases, bsAb includes the M428L / N434S half-life extension variant. The sequence of an exemplary TROP2×CD28 bsAb in the 1+1 Fab-scFv-Fc form is shown in Figure 30.

[0405] 3B:2+1 mAb-scFv format Another exemplary form utilizing the Fab domain and scFv is the 2+1 mAb-scFv form (schematically shown in Figure 15E), which includes a first monomer containing a first heavy chain covalently linked to a single-stranded Fv ("scFv") having a first antigen-binding specificity, i.e., VH-CH1-hinge-CH2-CH3-domain linker-scFv (where CH2-CH3 is the first heterodimer Fc domain), a second monomer containing the heavy chain, i.e., VH-CH1-hinge-CH2-CH3 (where CH2-CH3 is the second heterodimer Fc domain complementary to the first heterodimer Fc domain), and a light chain (LC) separately transfected so that a Fab domain having a second antigen-binding specificity is formed together with two VH domains. In particular, 1+1 Fab-scFv-Fc bsAb can utilize skeleton 1 or 11 in Figure 10. The skeleton utilizes the L368D / K370S (on HC):S364K / E357Q (on HC-scFv side) heterodimer Fc variant. HC further includes the pI variant N208D / Q295E / N384D / Q418E / N421D to increase the negative charge of the heavy chain. scFv utilizes a positively charged (GKPGS) 4-linker (SEQ ID NO: 24) between the VH and VL domains to increase the positive charge of the HC-scFv chain. In summary, these two approaches enable the easy purification of heterodimers from contaminated homodimers. The FcγR reduction variant used in this platform is the E233P / L234V / L235A / G236_ / S267K substitution in both the HC and scFv-Fc monomers. In some cases, bsAb includes the M428L / N434S half-life extended variant. An exemplary TROP2×CD28 bsAb sequence in the 2+1 mAb-scFv format is shown in Figure 31.

[0406] 3C: 1+1 Common Light Chain (CLC) type One common light chain form is the 1+1 common light chain (CLC) form (schematically shown in Figure 15A), which comprises a first monomer containing VH1-CH1-hinge-CH2-CH3, a second monomer containing VH2-CH1-hinge-CH2-CH3, and a third monomer containing VL-CL. VL pairs with VH1 to form a binding domain having a first antigen-binding specificity, and VL pairs with VH2 to form a binding domain having a second antigen-binding specificity. An exemplary sequence of TROP2×CD28 bsAb (based on the binding domains described herein) of the 1+1 CLC form is shown in Figure 38.

[0407] 3D: 2+1 Common Light Chain (CLC) format Another common light chain form is the 2+1 CLC form (schematically shown in Figure 15B), which comprises a first monomer containing VH1-CH1-linker-VH1-CH1-hinge-CH2-CH3, a second monomer containing VH2-CH1-hinge-CH2-CH3, and a third monomer containing VL-CL. VL pairs with the first and second VH1 to form a binding domain having a first antigen-binding specificity, and VL pairs with VH2 to form a binding domain having a second antigen-binding specificity. An exemplary sequence of TROP2×CD28 bsAb (based on the binding domains described herein) of the 2+1 CLC form is shown in Figure 39.

[0408] Example 3A provides a particularly useful scaffold for 1+1 Fab-scFv-Fc, which is also applicable to other forms. However, for the bispecific form, any number of heterodimerization approaches known in the art can be used in combination with any number of approaches for purifying heterodimers (including those shown in Figures 3-4) from contaminated homodimers. Any number of linkers known in the art can be used to ligate the VH and VL domains of scFv. Finally, it may be useful to maximize the serum half-life of bsAb, and any number of half-life extension variants known in the art can also be used for these bsAb. Despite the bsAb form, the CD28 bispecific antibody is monovalent for CD28 and incorporates an Fc variant that has been manipulated to eliminate FcγR binding to avoid potential superagonism. Such Fc variants include those shown in Figure 5.

[0409] Example 4: TROP2×CD28 bsAb induces cytokine secretion by T cells in the presence of a signaling pathway. 10,000 A431-β2M-null cells expressing αCD3 scFv (alternative to signal 1) (TROP2 高 The cells were plated overnight. Subsequently, A431 cells were treated with titrated doses of TROP2×CD28 bsAb or PBS control in the presence of purified T cells (E:T ratios of 10:1 and 1:1). One day after T cell seeding, IL2 secretion was measured using the MSD assay (Meso Scale Discovery, Rockville, Md.). The data shown in Figure 33 indicate that both TROP2×CD28 bsAb with monovalent and divalent TROP2 binding were active in inducing cytokine secretion.

[0410] Example 5: Combination with CD3 bsAb As described in Example 3, TROP2×CD28 bsAb is intended to be combined with CD3 bsAb. Such CD3 bsAb may utilize the CD3 binding domain shown in Figure 32. Furthermore, TROP2×CD28 bsAb can be combined with TROP2×CD3 bsAb (see, for example, U.S. Patent No. 9,382,329, U.S. Patent No. 10,245,321, and Liu H, Bai L, Huang L, et al. Bispecific antibody targeting TROP2xCD3 suppresses tumor growth of triple negative breast cancer. J Immunother Cancer. 2021;9(10):e003468.doi:10.1136 / jitc-2021-003468, incorporated herein by reference). They can also be combined with other CD3 bsAb.

[0411] XENP44424 and XENP44505, 1+1 Fab-scFv-Fc TROP2×CD28 bsAbs based on a novel 1F11 TROP2 binding domain (each possessing VHVL-oriented or VLVH-oriented CD28 scFv), were incubated with 1 μg / ml of exemplary B7H3×CD3 bsAb and a low effector:target ratio of 1:10 in the following cancer cell lines with different cell densities: 647-V (high TROP2), A431 (medium TROP2), MDA-MB-231 (low-medium TROP2), and DU145 (low-medium TROP2) to evaluate cytokine secretion by effector cells. The data shown in Figure 41 demonstrate that both TROP2×CD28 bsAbs dose-dependently and effectively combine with B7H3×CD3 bsAb to induce cytokine secretion. Notably, the orientation of CD3 scFv did not affect the activity. In a similar experiment, 1 μg / ml of XENP44424 or XENP44505 was incubated with exemplary B7H3×CD3 bsAb at titration doses. The data shown in Figure X are consistent with the data shown in Figure 41.

[0412] Example 6: Adjustment of TROP2×CD28 bsAb TROP2×CD28 bsAb was investigated in alternative forms, specifically in 1+1 CLC and 2+1 CLC configurations. In addition, alternative orientations of the second TROP2-binding domain in the 2+1 CLC configuration (i.e., stacked on top of the first TROP2-binding domain or the CD28-binding domain) were investigated. Furthermore, the effects of adjusting the binding affinity of the TROP2-binding domain were also investigated. Molecules embodying these various parameters were investigated in the following experiments.

[0413] In the first experiment, the titration dose of TROP2×CD28 bsAb was used, along with 1 μg / ml of exemplary B7H3×CD3 bsAb and OVCAR5(TROP2 高 The cells were incubated with pp65-MDA_MB-231 cancer cells (1:1 effector:target ratio). Data showing IL2 secretion after 24 hours are shown in Figures 42-43 and 50-51 (IFNγ data are not shown, but they show the same trend).

[0414] In the second experiment, a titration dose of TROP2×CD28 bsAb was incubated with 1 μg / ml of exemplary B7H3×CD3 bsAb, as well as DU145, MCF7, OvCAR5, and MDA-MB-231 cancer cells (1:1 effector:target ratio). Data showing IL2 secretion after 24 hours are shown in Figures 44-45. The IFNγ data shown in Figure 48 shows the same trend.

[0415] In summary, these data indicate that stronger TROP2 affinity variants exhibit higher potency (see, for example, comparison of XENP 44595, 44597, and 44599 in Figures 42, 44, and 50, which differ only in their TROP2 binding domains). Furthermore, no TROP2 avidity effect was observed in the 2+1 CLC configuration (see, for example, comparison of XENP44595 and XENP44601 in Figures 42-44). However, in the 2+1 CLC configuration, the orientation of the second TROP2 binding domain is important, and stacking the TROP2 binding domain on top of the CD28 binding domain, as in XENP44869 and XENP44871, reduced potency (see Figures 42 and 50). Notably, the control RSV×CD28 bsAb variants XENP44623 and XENP44624 were inactive.

[0416] Example 7: Combination with PD-1 blockade Checkpoint blockade (e.g., PD-1 blockade) offers broad utility in solid tumors and may be a useful therapeutic modality when combined with agonist antibody engagement of T cell costimulatory receptors on TILs to circumvent inhibition of the CD28 pathway by CTLA4. Therefore, we investigated the combination of the TROP2×CD28 bispecific antibody XENP44880 and XENP16432 (a bivalent anti-PD-1 mAb based on the variable region of nivolumab). Titrate doses of TROP2×CD28 bsAb were incubated with exemplary PD-1 mAb and MDA-MB-231 cancer cells (effector:target ratio of 10:1). IL2 secretion after 24 hours is shown in Figure 46. The TROP2×CD28 bsAb combination demonstrates effective combination with PD-1 blockade.

[0417] Example 8: T cell expansion and cytotoxicity 1 μg / ml of TROP2×CD28 bsAb XENP44599 was incubated with a co-culture of purified T cells and OVCAR5 cells, and with exemplary titration doses of B7H3×CD3 bsAb. T cell activation and proliferation, as well as tumor cell killing, are shown in Figure 47. This demonstrates that TROP2×CD28 improves T cell expansion and cytotoxicity of CD3 T cell engagers.

[0418] Example 9: Pharmacokinetics Cynomolgus monkeys were administered TROP2×CD28 bsAbs (XENP44599 and XENP44595, respectively) containing 2 nM or 13 nM TROP2-binding domains. As shown in Figure X, bsAbs with high TROP2 affinity were clearly rapidly metabolized by TMDD, while bsAbs with 13 nM TROP2 affinity showed a much better serum half-life.

[0419] Example 10: Selectivity TROP2 is highly expressed in cancer cells but is also expressed in normal cells. TROP2×CD28 bsAb is a 1+1 CLC form containing a 2, 5, or 13 nM TROP2-binding domain and a 21, 43, or 63 nM CD28-binding domain. OvCAR5(TROP2) 高 ) and OvCAR3(TROP2 低 The cells were incubated with T cells (1:1 effector:target ratio), 1 μg / ml exemplary EpCAM×CD3 bsAb, and a titration dose of TROP2×CD28 bsAb. The data are shown in Figures X and Y. 高 vs TROP2 低 This shows the changes in potency and efficacy in cell lines.

[0420] In another experiment, selected TROP2×CD28 bsAbs were incubated with exemplary EpCAM×CD3 bsAbs, purified T cells, and OVCAR5 cells to investigate cytotoxicity and T cell proliferation induced by CD28 bsAbs. The data shown in Figure 60 demonstrate that each CD28 bsAb dose-dependently induces target cell killing and T cell proliferation. The potency is consistent with the above experiment.

[0421] All cited references are explicitly incorporated herein by reference in their entirety.

[0422] While specific embodiments of the present invention are described above for illustrative purposes, it will be understood by those skilled in the art that numerous variations in detail can be made without departing from the present invention as described in the appended claims.

Claims

1. A heterodimer antibody, a) The first monomer, i) Single-stranded variable fragments (scFv); and ii) The first monomer comprising the first Fc domain, wherein the scFv is covalently bonded to the N-terminus of the first Fc domain using a domain linker; b) The second monomer comprising VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, wherein VH1 is a first variable weight domain and CH2-CH3 is a second Fc domain; and c) A light chain comprising VL1-CL from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain, The scFv includes a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2), The VH1 and VL1 together form a first antigen-binding domain (ABD), and the VH2 and VL2 together form a second ABD. The heterodimer antibody wherein one of the first ABD and the second ABD is a CD28-binding domain, and the other of the first ABD and the second ABD is a tumor-associated calcium signaling transducer 2 (TROP2)-binding domain.

2. The heterodimer antibody according to claim 1, wherein the scFv comprises a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

3. The heterodimer antibody according to claim 1, wherein the scFv comprises VL2-scFv linker-VH2 from the N-terminus to the C-terminus.

4. The heterodimer antibody according to any one of claims 1 to 3, wherein the first ABD is the TROP2 binding domain and the second ABD is the CD28 binding domain.

5. The heterodimer antibody according to claim 4, wherein VH1 and VL1 are selected from the following: (1) VH and VL or variants of any of the TROP2 binding domains in Figures 23-26 and 52-56; and (2) VH having the amino acid sequence of VH or a variant thereof as shown in Figure 36, and VL having the amino acid sequence of the 1F11-1A3.315 L1 variable light domain (see Figure 36A) or a variant thereof.

6. The heterodimer antibody according to claim 4 or 5, wherein VH2 and VL2 are selected from one of the following: (1) VH and VL of any of the CD28 binding domains in Figures 16, 19, and 22, or a variant thereof, (2) (i) VH or a variant thereof from Figure 16, 17, or 34, and (ii) VL or a variant thereof from Figure 16, 18, or 36.

7. The heterodimer antibody according to any one of claims 1 to 6, wherein the first Fc domain and the second Fc domain are each variant Fc domains.

8. The heterodimer antibody according to claim 7, wherein the first and second Fc domains comprise a set of heterodimerized scuba arians selected from the following heterodimerized variants: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, the numbering according to EU numbering.

9. The heterodimer antibody according to claim 8, wherein the first and second Fc domains comprise the heterodimerized scuba rianto S364K / E357Q:L368D / K370S, and the numbering follows EU numbering.

10. A heterodimer antibody according to any one of claims 7 to 9, wherein each of the first and second Fc domains comprises one or more attenuation variants.

11. The heterodimer antibody according to claim 10, wherein one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, and the numbering follows EU numbering.

12. A heterodimer antibody according to any one of claims 7 to 11, wherein one of the first or second monomers further comprises one or more pI variants.

13. The heterodimer antibody according to claim 12, wherein the CH1-hinge-CH2-CH3 of the second monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, and the numbering follows EU numbering.

14. The CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, The first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, The heterodimer antibody according to any one of claims 7 to 13, wherein the numbering follows EU numbering.

15. The heterodimer antibody according to any one of claims 8 to 14, wherein the first and second variant Fc domains each further comprise the amino acid variant 428L / 434S.

16. The heterodimer antibody according to any one of claims 1 to 15, wherein the scFv linker is GKPGSGKPGSGKPGSGKPGS (Sequence ID 24).

17. A heterodimer antibody, a) A first monomer comprising VH1-CH1-first domain linker-scFv-second domain linker-CH2-CH3 from the N-terminus to the C-terminus, The first monomer, wherein VH1 is a first variable weight domain and CH2-CH3 is a first Fc domain; b) The second monomer comprising VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, wherein CH2-CH3 is a second Fc domain; and c) A first light chain comprising VL1-CL from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain; and d) A second light chain comprising VL1-CL extending from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain, The scFv includes a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2), The VH1 of the first monomer and the VL1 of the first light chain and the VH1 of the second monomer and the VL1 of the second light chain each form a first antigen-binding domain (ABD), and the VH2 and VL2 form a second ABD. The heterodimer antibody wherein one of the first ABD and the second ABD is a CD28-binding domain, and the other of the first ABD and the second ABD is a mesothelin (TROP2)-binding domain.

18. The heterodimer antibody according to claim 17, wherein the scFv comprises a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

19. The heterodimer antibody according to claim 17, wherein the scFv comprises VL2-scFv linker-VH2 from the N-terminus to the C-terminus.

20. The heterodimer antibody according to any one of claims 17 to 19, wherein the first ABD is the TROP2 binding domain and the second ABD is the CD28 binding domain.

21. The heterodimer antibody according to claim 20, wherein VH1 and VL1 are either VH and VL or a variant thereof of the TROP2 binding domains shown in Figures 23-26 and 52-56.

22. The heterodimer antibody according to claim 20 or 21, wherein VH2 and VL2 are selected from one of the following: (1) VH and VL of any of the CD28 binding domains in Figures 16, 19, and 22, or a variant thereof, (2) (i) VH or a variant thereof from Figure 16, 17, or 34, and (ii) VL or a variant thereof from Figure 16, 18, or 36.

23. The heterodimer antibody according to any one of claims 17 to 22, wherein the first Fc domain and the second Fc domain are each variant Fc domains.

24. The heterodimer antibody according to claim 23, wherein the first and second Fc domains comprise a set of heterodimerized scuba arians selected from the following heterodimerized variants: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, the numbering according to EU numbering.

25. The heterodimer antibody according to claim 24, wherein the first and second Fc domains comprise the heterodimerized scuba rianto S364K / E357Q:L368D / K370S, and the numbering follows EU numbering.

26. A heterodimer antibody according to any one of claims 23 to 25, wherein each of the first and second Fc domains comprises one or more attenuation variants.

27. The heterodimer antibody according to claim 26, wherein one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, and the numbering follows EU numbering.

28. A heterodimer antibody according to any one of claims 23 to 27, wherein one of the first or second monomers further comprises one or more pI variants.

29. The heterodimer antibody according to claim 28, wherein the CH1-hinge-CH2-CH3 of the second monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, and the numbering follows EU numbering.

30. The CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, The first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, The heterodimer antibody according to any one of claims 23 to 29, wherein the numbering follows EU numbering.

31. The heterodimer antibody according to any one of claims 24 to 30, wherein the first and second variant Fc domains each further comprise the amino acid variant 428L / 434S.

32. The heterodimer antibody according to any one of claims 17 to 31, wherein the scFv linker is GKPGSGKPGSGKPGSGKPGS (Sequence ID 24).

33. A heterodimer antibody, a) A first monomer comprising VH1-CH1-hinge-CH2-CH3-domain linker-scFv from the N-terminus to the C-terminus, The first monomer, wherein VH1 is a first variable weight domain and CH2-CH3 is a first Fc domain; b) The second monomer comprising VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, wherein VH1 is a first variable weight domain and CH2-CH3 is a second Fc domain; and c) A first light chain comprising VL1-CL from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain; and d) A second light chain comprising VL1-CL extending from the N-terminus to the C-terminus, wherein VL1 is a first variable light domain and CL is a constant light domain, The scFv includes a second VH domain (VH2), an scFv linker, and a second variable light domain (VL2), The VH1 of the first monomer and the VL1 of the first light chain and the VH1 of the second monomer and the VL1 of the second light chain each form a first antigen-binding domain (ABD), and the VH2 and VL2 form a second ABD. The heterodimer antibody wherein one of the first and second ABDs binds to human CD28, and the other of the first and second ABDs binds to TROP2.

34. The heterodimer antibody according to claim 33, wherein the scFv comprises a VH2-scFv linker-VL2 from the N-terminus to the C-terminus.

35. The heterodimer antibody according to claim 33, wherein the scFv comprises VL2-scFv linker-VH2 from the N-terminus to the C-terminus.

36. The heterodimer antibody according to any one of claims 33 to 35, wherein the first ABD is the TROP2 binding domain and the second ABD is the CD28 binding domain.

37. The heterodimer antibody according to claim 36, wherein VH1 and VL1 are either VH and VL or variants of the TROP2 binding domains shown in Figures 23-26 and 52-56.

38. The heterodimer antibody according to claim 36 or 37, wherein VH2 and VL2 are selected from one of the following: (1) VH and VL of any of the CD28 binding domains in Figures 16, 19, and 22, or a variant thereof, (2) (i) VH or a variant thereof from Figure 16, 17, or 34, and (ii) VL or a variant thereof from Figure 16, 18, or 36.

39. The heterodimer antibody according to any one of claims 33 to 38, wherein the first Fc domain and the second Fc domain are each variant Fc domains.

40. The heterodimer antibody according to claim 39, wherein the first and second Fc domains comprise a set of heterodimerized scuba arians selected from the following heterodimerized variants: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, the numbering according to EU numbering.

41. The heterodimer antibody according to claim 40, wherein the first and second Fc domains comprise the heterodimerized scuba rianto S364K / E357Q:L368D / K370S, and the numbering follows EU numbering.

42. A heterodimer antibody according to any one of claims 39 to 41, wherein each of the first and second Fc domains comprises one or more attenuation variants.

43. The heterodimer antibody according to claim 42, wherein one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, and the numbering follows EU numbering.

44. A heterodimer antibody according to any one of claims 39 to 43, wherein one of the first or second monomers further comprises one or more pI variants.

45. The heterodimer antibody according to claim 44, wherein the CH1-hinge-CH2-CH3 of the second monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, and the numbering follows EU numbering.

46. The CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, The first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, The heterodimer antibody according to any one of claims 39 to 45, wherein the numbering follows EU numbering.

47. The heterodimer antibody according to any one of claims 40 to 46, wherein the first and second variant Fc domains each further comprise the amino acid variant 428L / 434S.

48. The heterodimer antibody according to any one of claims 33 to 47, wherein the scFv linker is GKPGSGKPGSGKPGSGKPGS (Sequence ID 24).

49. a) i) a TROP2 binding domain including a first variable heavy domain (VH1) and ii) a first variable light domain (VL1), b) A bispecific antibody comprising an anti-CD28 binding domain containing i) a second variable heavy domain (VH2) and ii) a second variable light domain (VL2).

50. The bispecific antibody according to claim 49, wherein VH1 and VL1 are either VH and VL or variants of the TROP2 binding domains shown in Figures 23-26 and 52-56.

51. The bispecific antibody according to claim 49 or 50, wherein VH2 and VL2 are selected from one of the following: (1) VH and VL of any of the CD28 binding domains in Figures 16, 19, and 23, or their variants, (2) (i) VH or a variant thereof as shown in Figure 17 and (ii) VL or a variant thereof as shown in Figure 18.

52. The bispecific antibody according to any one of claims 49 to 51, wherein the bispecific antibody further comprises a first Fc domain and a second Fc domain.

53. The bispecific antibody according to claim 52, wherein the first and second Fc domains comprise a set of heterodimerized scuba ariants selected from the following heterodimerized variants: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, the numbering according to EU numbering.

54. The bispecific antibody according to claim 53, wherein the first and second Fc domains comprise the heterodimerized scuba rianto S364K / E357Q:L368D / K370S, and the numbering follows EU numbering.

55. The bispecific antibody according to any one of claims 52 to 54, wherein each of the first and second Fc domains comprises one or more attenuation variants.

56. The bispecific antibody according to claim 55, wherein one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, and the numbering follows EU numbering.

57. The bispecific antibody according to any one of claims 52 to 56, wherein one of the first or second monomers further comprises one or more pI variants.

58. The bispecific antibody according to claim 57, wherein the pI variants N208D / Q295E / N384D / Q418E / N421D, numbered according to EU numbering.

59. a) A first nucleic acid encoding a first monomer according to any one of claims 1 to 48; b) A second nucleic acid encoding the second monomer according to any one of claims 1 to 48; and c) A nucleic acid composition comprising a light chain according to any one of claims 1 to 16 or a third nucleic acid encoding the first and second light chains according to any one of claims 17 to 48.

60. a) A first expression vector comprising the first nucleic acid according to claim 59; b) A second expression vector comprising the second nucleic acid described in claim 59; and c) An expression vector composition comprising a third expression vector each containing the third nucleic acid described in claim 59.

61. A host cell comprising the expression vector composition described in claim 60.

62. A method for producing a heterodimer antibody according to any one of claims 1 to 48, comprising culturing a host cell according to claim 61 under conditions in which the heterodimer antibody is expressed, and recovering the heterodimer antibody.

63. A method for treating a patient requiring treatment for a TROP2-related cancer, comprising administering to the patient a heterodimer antibody according to any one of claims 1 to 48.

64. A method for treating a patient requiring treatment for a TROP2-related cancer, comprising administering to the patient a heterodimer antibody and an anti-CD3 × anti-TROP2 bispecific antibody according to any one of claims 1 to 48.

65. A method for treating a patient requiring treatment for a TROP2-related cancer, comprising administering to the patient a bispecific antibody according to any one of claims 49 to 58.

66. A method for treating a patient requiring treatment for a TROP2-related cancer, comprising administering to the patient a bispecific antibody according to any one of claims 49 to 58 and an anti-CD3 × anti-TROP2 bispecific antibody.

67. It is a polyvalent antibody, a) A first monomer comprising VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, wherein VH1 is a first variable weight domain and CH2-CH3 is a first Fc domain; b) The second monomer comprising VH2-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, wherein VH2 is a second variable weight domain and CH2-CH3 is a second Fc domain; c) A first common light chain and a second common light chain, each containing VL and CL, wherein VL is a variable domain and CL is a constant light domain, comprising the first common light chain and the second common light chain, The VH1 and VL of the first common light chain form a first antigen-binding domain (ABD), and the VH2 and VL of the second common light chain together form a second ABD. The first common light chain and the second common light chain have the same amino acid sequence. The polyvalent antibody wherein one of the first ABD and the second ABD is a CD28-binding domain, and the other of the first ABD and the second ABD is a tumor-associated calcium signaling transducer 2 (TROP2)-binding domain.

68. The polyvalent antibody according to claim 67, wherein the first ABD is the TROP2 binding domain and the second ABD is the CD28 binding domain.

69. The polyvalent antibody according to claim 68, wherein VH1 is the variable weight domain shown in Figure 36 or 52-56, VH2 is the variable weight domain shown in Figure 34, and the VL of the first and second common light chains is the 1F11-1A3.315[TROP2]_L1 variable light domain (see Figure 36A).

70. The polyvalent antibody according to any one of claims 67 to 69, wherein the first Fc domain and the second Fc domain are each variant Fc domains.

71. The polyvalent antibody according to claim 70, wherein the first and second Fc domains comprise a set of heterodimerized scuba ariants selected from the following heterodimerized variants: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, the numbering according to EU numbering.

72. The polyvalent antibody according to claim 71, wherein the first and second Fc domains comprise the heterodimerized scuba rianto S364K / E357Q:L368D / K370S, and the numbering follows EU numbering.

73. The polyvalent antibody according to any one of claims 70 to 72, wherein each of the first and second Fc domains comprises one or more attenuation variants.

74. The polyvalent antibody according to claim 73, wherein one or more attenuation variants include E233P / L234V / L235A / G236del / S267K, and the numbering follows EU numbering.

75. The polyvalent antibody according to any one of claims 71 to 74, wherein one of the first or second monomers further comprises one or more pI variants.

76. The polyvalent antibody according to claim 75, wherein the CH1-hinge-CH2-CH3 of the second monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, and the numbering follows EU numbering.

77. The CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, The first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, The numbering follows EU numbering, and the multivalent antibody is as described in any one of claims 70 to 76.

78. The heterodimer antibody according to any one of claims 71 to 77, wherein the first and second variant Fc domains each further comprise the amino acid variant 428L / 434S.

79. It is a polyvalent antibody, a) The first monomer comprising VH1-CH1-linker-VH1-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, wherein each VH1 is a first variable weight domain and CH2-CH3 is a first Fc domain; b) The second monomer comprising VH2-CH1-hinge-CH2-CH3 from the N-terminus to the C-terminus, wherein VH2 is a second variable weight domain and CH2-CH3 is a second Fc domain; c) A first common light chain, a second common light chain, and a third common light chain, each containing a VL-CL, wherein VL is a variable domain and CL is a constant light domain, the first common light chain, the second common light chain, and the third common light chain, Each of the VH1 molecules of the first monomer pairs with the VL of the first common light chain or the second common light chain to form two first antigen-binding domains (ABDs), and the VH2 and VL of the third common light chain together form a second ABD. The first common light chain, the second common light chain, and the third common light chain each have the same amino acid sequence. The polyvalent antibody wherein the first ABD is a CD28-binding domain and the second ABD is a tumor-associated calcium signaling transducer 2 (TROP2)-binding domain, or the first ABD is a TROP2-binding domain and the second ABD is a CD28-binding domain.

80. The polyvalent antibody according to claim 79, wherein the first ABD is a TROP2 binding domain and the second ABD is a CD28 binding domain.

81. The polyvalent antibody according to claim 80, wherein VH1 is the variable weight domain shown in Figure 36, VH2 is the variable weight domain shown in Figures 34 or 52-56, and the VL of the common light chain of the first, second, and third is the 1F11-1A3.315[TROP2]_L1 variable light domain (see Figure 36A).

82. The polyvalent antibody according to any one of claims 79 to 81, wherein the first Fc domain and the second Fc domain are each variant Fc domains.

83. The polyvalent antibody according to claim 82, wherein the first and second Fc domains comprise a set of heterodimerized scuba ariants selected from the following heterodimerized variants: S364K / E357Q:L368D / K370S; S364K:L368D / K370S; S364K:L368E / K370S; D401K:T411E / K360E / Q362E; and T366W:T366S / L368A / Y407V, the numbering according to EU numbering.

84. The polyvalent antibody according to claim 83, wherein the first and second Fc domains comprise the heterodimerized scuba rianto S364K / E357Q:L368D / K370S, and the numbering follows EU numbering.

85. The polyvalent antibody according to any one of claims 82 to 84, wherein each of the first and second Fc domains comprises one or more attenuation variants.

86. The polyvalent antibody according to claim 85, wherein one or more of the diminished variants include E233P / L234V / L235A / G236del / S267K, and the numbering follows EU numbering.

87. The polyvalent antibody according to any one of claims 83 to 86, wherein one of the first or second monomers further comprises one or more pI variants.

88. The polyvalent antibody according to claim 87, wherein the CH1-hinge-CH2-CH3 of the second monomer contains the pI variant N208D / Q295E / N384D / Q418E / N421D, and the numbering follows EU numbering.

89. The CH1-hinge-CH2-CH3 of the second monomer contains the amino acid variant E233P / L234V / L235A / G236del / S267K / L368D / K370S / N208D / Q295E / N384D / Q418E / N421D, The first Fc domain contains the amino acid variant E233P / L234V / L235A / G236del / S267K / S364K / E357Q, The numbering follows EU numbering, and the multivalent antibody is as described in any one of claims 82 to 88.

90. The polyvalent antibody according to any one of claims 83 to 89, wherein the first and second variant Fc domains each further comprise the amino acid variant 428L / 434S.

91. a) A first nucleic acid encoding a first monomer according to any one of claims 67 to 90; b) A second nucleic acid encoding the second monomer according to any one of claims 67 to 90; and c) A nucleic acid composition comprising a third nucleic acid encoding a common light chain as described in any of claims 67 to 90.

92. a) A first expression vector comprising the first nucleic acid according to claim 91; b) A second expression vector comprising the second nucleic acid described in claim 91; and c) An expression vector composition comprising a third expression vector each containing the third nucleic acid described in claim 91.

93. A host cell comprising the expression vector composition described in claim 92.

94. A method for producing a polyvalent antibody according to any one of claims 67 to 90, comprising culturing a host cell according to claim 93 under conditions in which the heterodimeric antibody is expressed, and recovering the heterodimeric antibody.

95. A method for treating a TROP2-related cancer in a patient requiring treatment for the TROP2-related cancer, comprising administering a heterodimer antibody according to any one of claims 67 to 90 to the patient.

96. A method for treating a patient requiring treatment for a TROP2-related cancer, comprising administering to the patient a heterodimer antibody and an anti-CD3 × anti-TROP2 bispecific antibody according to any one of claims 67 to 90.

97. A TROP-2 binding construct including variable weight domains and variable light domains selected from the following: (1) Any VH and VL of the TROP2 binding domain in Figures 23-26 and 52-56, or a variant thereof; and (2) VH or its variant as shown in Figure 36, and 1F11-1A3.315L1VL (see Figure 36A) or its variant.

Images